After a mysterious banner for Tony Hawks Pro Skater appeared in Call of Duty: Black Ops 6, a countdown site has cropped up. Like the banner, its counting down to something on March 4th, specifically 11:30 AM PST.There are a few more clues to infer from this alongside the Activision logo at the bottom is Iron Galaxy Studios, which could be working on this new project. The studio co-developed the PC version of Tony Hawks Pro Skater 1 + 2 alongside Vicarious Vision. The websites address also mentions Chicago, potentially indicating the levels inclusion (following its remastered appearance in the last title).Rumors have circulated about the countdown being for Tony Hawks Pro Skater 3 + 4, a remastered compilation of the third and fourth games in the series. Pro skater Tyshawn Jones also revealed hes in a new Tony Hawk remaster, though he didnt specify if it was for the Pro Skater series. Stay tuned for more details and updates in the coming weeks.

Arrowhead Game Studios is apparently working on a new collaboration event for Helldivers 2. As caught by GamesRadar, the studio revealed through the games official Discord channel that it is working on something new when asked by a player about future crossover events.A developer that goes by the handle Sephez revealed on Discord that Arrowhead is in the very early stages of something new.The reveal was in response to a Helldivers 2 players saying that a potential crossover between the co-op shooter along with another co-op shooter, Deep Rock Galactic, would be interesting.While the developer was careful to not reveal any details, the use of the word new would indicate that it wont be a crossover event similar to what weve seen before.The last major crossover event we saw in Helldivers 2 was with the (these days ignored) FPS franchise Killzone. The crossover primarily focused on Killzone 2, and went live back in December. As part of the event, the game got a host of new content, including new armour sets, weapons, and cosmetic items like banners and capes.While the crossover was successful, it was a source of some controversy owing to high price tags for the new items introduced to the game. The studio took to Twitter to respond to these criticisms.As this was our first crossover offering, weve been listening to the response, and the feedback weve received from our community surveys is that crossover content is really high up on your wishlists, said the studio. At the same time, some of you are not happy with the prices of the items, and we will revisit this.The most recent bit of content released for Helldivers 2 was the Servants of Freedom Premium Warbond, which brought with it weapons themed around explosives, such as the LAS-17 Double-Edge Sickle and the GP-20 Ultimatum. The update also brought with it the B-100 Portable Hellbomb, which can be used to clear out buildings.Helldivers 2 is around a year old at this point, and Arrowhead Game Studios decided to celebrate the event with a look back at some of its favourite moments from the Galactic War campaign. Featuring the mysterious game master Joel, a video took us on a trip down memory lane by giving us a look back at the legendary Defense of Mort.Helldivers 2 creative director Johan Pilestedt announced last month that he would be going on a sabbatical. In his announcement, Pilestedt spoke about needing a break after having worked on the franchise non-stop since 2013. He kicked off his sabbatical by hosting a game jam along with senior developers at Arrowhead.While Pilestedt is on his sabbatical, studio CEO Shams Jorjani spoke about how it would keep working on Helldivers 2, since development of the game has been a team effort. Jorjani spoke about how the game is currently being headed up by Mikael Eriksson, and how its future isnt dependent on a singular Jesus figure.Helldivers 2 is available on PC and PS5. Check out our review for more details.

After a public callout, the developers of Hades took to social media to clarify that they are not intending to recast union voice actors.Read More

Sony has pushed back against claims levied by former Bungie director Christopher Barrett, who sued the two companies in 2024 after he was fired from the Destiny 2 and Marathon developer. The PlayStation maker released a 128-page response (first obtained by GameFile), looking to dismiss six of the seven counts in Barrett's lawsuit.At the time, Barrett was laid off for alleged misconduct, which concerned "behaving inappropriately" toward at least eight women in different departments at Bungie. Barrett argued his firing was "unfounded" and accused Bungie and Sony of firing him to avoid giving him a $45 million payout from his employee agreement.Sony's response counters Barrett's accusations, saying the investigation against him revealed his "consistent" behavior toward the unnamed employees."Hetargetedalower-level, femaleemployeehewasnt working with directly, initially engaged in friendly conversation, and progressively pushed boundaries with theemployeeby making subtle references to her physical appearanceor expressinghis interestin her romantically," the response begins. "Barrett attempted to createan unprofessional level ofintimacy withhisvictims."The response goes on to allege Barrett frequently brought up "inappropriate topics" toward these women, such as their physical appearance, and would text them "at all hours of the day and night." Several texts are cited in the response, and were further corroborated by Bloomberg's Jason Schreier, who first reported Barrett's firing.Barrett and Bungie's alleged relationshipAccording to Sony, several of the victims told Barrett to stop or reported him to Bungie's HR department. However, it's claimed he would continue the pattern after initially backing off. His alleged behavior "made multiple female employees who were subordinate to himfeeluncomfortable, fear retaliation, andfeelvictimizedby their encounters withhim.Thisimpeded their ability to perform their job duties and thereby hindered [our] business operations."Legal representatives for Barrett told Game File that Sony continues to "disingenuously cherrypick text messages and alleged conversations and make unsupported and conclusory statements to defame Christopher and justify terminating him to avoid paying him what he was owed under his employment agreements. [...] Nothing in Sonys response provides a legitimate legal or factual basis to terminate Christopher for cause."Barret has previously argued Bungie had a history of workplace toxicity that he was witness to. His initial lawsuit mentioned a "very senior executive" texting him and other high-level staffers inappropriate material and texts about the appearance of women employees at the studio. Sony said it "lacks knowledge or information sufficient to form a belief as to the truth of the allegations," but remain firm in its overall argument."Barretts allegations that hewas'scapegoated,' 'falsely accused, and 'defamed' are outrageous considering the evidence," continued Sony. "Hundreds ofBarretts own written statements and thedirect testimony ofhis many victims will demonstrate his termination for Cause was justified and this lawsuit is meritless."Sony's full response can be read here.

The General Services Administration (GSA), which manages buildings owned by the federal government, is planning to shut down all of its electric vehicle chargers nationwide, describing them as not mission critical. The agency, which manages contracts for the governments vehicle fleets, is also looking to offload newly purchased EVs. The GSA currently operates several hundred EV chargers across the country, with approximately 8,000 plugs that are available for government-owned EVs as well as federal employees personally owned vehicles. The official guidance instructing federal workers to begin the process of shutting down the chargers will be announced internally next week, according to a source with knowledge of the plans. Some regional offices have been told to start taking their chargers offline, according to an email viewed by The Verge. Are you a current or former GSA worker? Reach out securely on a personal device with tips to Andrew Hawkins via Signal at andrewhawkins.35.As GSA has worked to align with the current administration, we have received direction that all GSA owned charging stations are not mission critical, the email reads. The GSA is working on the timing of canceling current network contracts that keep the EV chargers operational. Once those contracts are canceled, the stations will be taken out of service and turned off at the breaker, the email reads. Other chargers will be turned off starting next week. Neither Government Owned Vehicles nor Privately Owned Vehicles will be able to charge at these charging stations once theyre out of service, it concludes.At the GSAs Denver office, employees were told that EV chargers at four federally owned buildings would be taken offline next week. The news was first reported by Colorado Public Radio. Under the Biden administration, the GSA was in charge of implementing the presidents plan to phase out the federal governments use of gas-powered vehicles in favor of EVs. The federal government owns approximately 650,000 vehicles, more than half of which were to be replaced with EVs.Those new EVs would need reliable places to charge. Former President Joe Bidens signature climate legislation, the Inflation Reduction Act, included $975 million for the GSA to upgrade federal buildings across the country with emerging and sustainable technologies. The aim was to achieve a net-zero emissions federal building portfolio by 2045, which included EV chargers.According to a March 2024 update, the GSA had ordered over 58,000 EVs and begun installing more than 25,000 charging ports, adding to the 8,000 already in use across the government. An interactive map showing the location of all GSA-owned chargers has been taken offline as of February this year. (An older version is available through the Wayback Machine.)The GSA will also begin offloading the EVs it purchased under the Biden administration, the source said. Its unclear whether those vehicles will be sold or simply put away in storage. Its also unclear whether other federal agencies will be making similar decisions for their own EVs, although many of those agencies tend to use the GSAs EV chargers for their own plug-in vehicles. Neither Government Owned Vehicles nor Privately Owned Vehicles will be able to charge at these charging stations once theyre out of service.President Donald Trump campaigned on a promise to roll back his predecessors EV policies, which he falsely labeled a mandate. And since his inauguration, he halted a $5 billion program to install new public EV chargers across the country, signed an executive order rescinding Bidens directives to purchase new EVs for the federal governments vehicle fleet, and signaled his intention to eliminate the federal EV tax credit and other incentives for consumers.Unlike gas-powered vehicles, EVs generate no tailpipe pollution. Burning fossil fuels like gasoline and diesel release carbon dioxide, a greenhouse gas, into the environment. These emissions have been proven to cause climate change, which supercharges extreme weather events like wildfires, hurricanes, and flooding. Transportation, including personal vehicle usage, accounts for about 28 percent of all US greenhouse gas emissions, according to the Environmental Protection Agency.A spokesperson for the GSA did not immediately respond to a request for comment. The GSA also plans on selling approximately 500 buildings as part of the Trump administrations efforts to gut the federal government, Wired reports. With additional reporting by Mia SatoSee More:

While LLMs have shown remarkable advancements in general-purpose applications, their development for specialized fields like medicine remains limited. The complexity of medical knowledge and the scarcity of high-quality, domain-specific data make creating highly efficient medical LLMs challenging. Although models like GPT-4 and DeepseekR1 have demonstrated impressive capabilities across industries, their adaptation to the medical domain is hindered by the intricate nature of medical terminology, diverse disciplines, and constantly evolving literature. Unlike general applications, medical AI must interpret highly technical language and provide precise, contextually relevant responses, which traditional LLMs struggle to achieve.One major obstacle in building effective medical LLMs is the limited accessibility of high-quality training data, which is restricted due to privacy concerns and regulatory barriers. Medical datasets consist of structured and unstructured information, including clinical notes, textbooks, and research articles, making comprehensive model training difficult. While approaches like fine-tuning general LLMs on medical datasets and applying transfer learning have been explored, these methods often fail to grasp the depth of medical knowledge fully. As a result, such models may perform well on specific tasks but lack the nuanced understanding necessary for complex medical inquiries, highlighting the need for more refined training strategies.Researchers at Baichuan Inc. introduced Baichuan-M1, a specialized large language model series designed specifically for medical applications. Unlike traditional models that refine existing architectures through additional pretraining or post-training, Baichuan-M1 is built from scratch with a strong focus on medical expertise. Trained on 20 trillion tokens, including both general and medical-specific data, the model balances broad language understanding with domain-specific precision. It excels in general tasks like coding and mathematics and in medical applications such as diagnostics and treatment recommendations. With an optimized Transformer architecture, Baichuan-M1 sets a new benchmark for AI-driven advancements in healthcare.The model architecture follows Llama and similar frameworks, incorporating pre-norm RMSNorm, SwishGlu in the FFN layer, and rotary position embeddings. The study integrates global and sliding window attention to optimize inference efficiency, increasing the head dimension to 256 for global layers. Additionally, temporal short convolutions on key-value attention enhance in-context learning. The model employs a hybrid tokenizer for medical and general text, a curriculum-based training strategy with progressive data complexity, and adaptive gradient clipping for stability. Supervised fine-tuning refines general reasoning and medical-specific tasks, ensuring robust language understanding, medical reasoning, and long-document handling capabilities while maintaining inference efficiency.Using various benchmarks, baichuan-M1-14B-Bases code and mathematical abilities were evaluated against the Qwen2.5 series models. Code generation performance was tested with the EvalPlus framework and Bigcodebench, while mathematical proficiency was assessed using MATH and CMATH datasets. Although the 14B-Instruct variant still lags behind proprietary models like Claude-3.5-Sonnet and GPT-4o, the gap has narrowed significantly. The results demonstrate that Baichuan-M1-14B-Base performs competitively in certain tasks, showcasing its code generation and mathematical reasoning strengths compared to other advanced models.In conclusion, Traditional methods for adapting LLMs to specialized fields often involve fine-tuning existing models. However, experiments suggest that further training on pre-existing models can hinder domain-specific improvements without sacrificing general performance. In the medical domain, fine-tuning general models with domain-specific data may be less effective than training from scratch. Baichuan-M1 was developed with this approach, using 20 trillion tokens to enhance medical expertise while maintaining general capabilities. Open-sourcing Baichuan-M1-14B allows further research, though challenges remain in rare disease diagnosis and real-world applications. Its continued evolution could significantly advance AI-driven medical decision-making.Check outBaichuan-M1-14B-Instruct.All credit for this research goes to the researchers of this project. Also,feel free to follow us onTwitterand dont forget to join our75k+ ML SubReddit. Sana HassanSana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions.Sana Hassanhttps://www.marktechpost.com/author/sana-hassan/xAI Releases Grok 3 Beta: A Super Advanced AI Model Blending Strong Reasoning with Extensive Pretraining KnowledgeSana Hassanhttps://www.marktechpost.com/author/sana-hassan/Learning Intuitive Physics: Advancing AI Through Predictive Representation ModelsSana Hassanhttps://www.marktechpost.com/author/sana-hassan/Microsoft AI Releases OmniParser V2: An AI Tool that Turns Any LLM into a Computer Use AgentSana Hassanhttps://www.marktechpost.com/author/sana-hassan/Enhancing Diffusion Models: The Role of Sparsity and Regularization in Efficient Generative AI

Arbitration for AI: A New Frontier in Governing Uncensored Models 0 like February 21, 2025Share this postAuthor(s): Mohit Sewak, Ph.D. Originally published on Towards AI. AI Justice League: When models go wild, arbitration calls for backup!Hey there, future AI whisperers and digital dynamos! Dr. Sewak here, your friendly neighborhood AI researcher, and today we are diving deep into a topic thats hotter than Bengaluru traffic in summer Uncensored AI. Now, before you picture robots running wild and taking over the world like in some Hollywood blockbuster Terminator anyone? lets pump the brakes a bit. Its not quite Skynet just yet, but its definitely a space where things can get interesting.Think of AI models like kids in a candy store. Censored or aligned AI? Those are the well-behaved kids, the ones told No candy before dinner! (OpenAI, 2024a). Theyre trained to be polite, avoid saying naughty words, and generally play nice. Theyve got filters stricter than my mom checking my search history back in the day. These models are all about ethical guidelines, American values, and political correctness basically, theyre trying to be the golden children of AI (JarvisLabs, 2024).Pro Tip: Always question the source, even if its AI. Uncensored doesnt mean unbiased or factual it just means unfiltered!But then, BAM! Enter the uncensored models (Zem, 2024). These are the rebels, the rule-breakers, the AI equivalent of that cool kid in school who wore ripped jeans and listened to rock music. Theyre designed to give it to you straight, no chaser, no filters. Theyre like, You want info? Ill give you ALL the info, the good, the bad, and the downright weird! They aim to process and spit out everything theyve learned, holding nothing back. Sounds exciting, right? Like finally, AI that tells it like it is!Trivia Time: Did you know the term censorship in AI is kinda borrowed from human societies? But whos the censor for AI? Thats the million-dollar question were tackling!With great power, comes great responsibility. Uncle Ben, Spider-Man.(Yeah, even Spidey knew AI ethics before it was cool. Dr.Sewak)Now, as someone whos been in the AI trenches at places like Google, NVIDIA, Microsoft R&D, IBM Labs, and even wrestled with AI in the banking world at Bank of New York Mellon, I can tell you firsthand this uncensored thing? Its a double-edged sword. Its got the potential to unlock crazy innovation think research breakthroughs, wild creativity, and access to info like never before. But, and this is a BIG but, it also opens up a Pandoras Box of risks (Bommasani et al., 2021; Weidinger et al., 2021). Were talking misinformation on steroids, content so unethical itd make a politician blush, and outputs that could be weaponized faster than you can say deepfake (Metz, 2023; Solaiman et al., 2023).Uncensored AI: Pandoras Box or Treasure Chest? Handle with extreme care!So, what do we do? Do we slap a giant CENSORED sticker on all AI and call it a day? Nah, thats like putting toothpaste back in the tube messy and impossible. Plus, as someone whos authored books like Deep Reinforcement Learning (Springer Nature) and Convolutional Neural Networks, (Packt), I believe in pushing boundaries, not building walls. Censorship can be a slippery slope, leading to over-alignment, where AI becomes so vanilla, its about as useful as a chocolate teapot (Digit, 2025). And trust me, in the AI world, being useless is the ultimate insult.Thats where arbitration comes in, my friends. Think of it as the AI referee, the wise old judge, the okay, you get it. Its a way to manage the chaos, to set some ground rules for these uncensored models without killing their innovative spirit. Its about finding that sweet spot, that balance between wild west freedom and responsible AI citizenship. Its a New Frontier, baby! And were gonna explore it together. Ready to ride? Lets dive into the nitty-gritty of this arbitration thing and see if we can tame this uncensored AI beast!The Uncensored AI Beast: Taming the Wild West with ArbitrationCan we harness the power of unrestricted AI without unleashing chaos? A novel framework for responsible governance.Alright, so weve established that uncensored AI is like that super-talented but slightly reckless friend we all have. They can do amazing things, but you also kinda hold your breath when theyre in charge of the party playlist. The core issue boils down to this censorship thing or rather, the lack of it (Ovadia, 2023).Pro Tip: Think of censorship in AI not as bad, but as alignment. Its about aligning AI with our values, not just silencing it.The only way to do great work is to love what you do. Steve Jobs.(And in AI, great work means responsible work, even with uncensored models. Dr.Sewak)Now, censored AI or as the cool kids call it, aligned AI is all about filters, baby (OpenAI, 2024b). Think of Instagram filters, but instead of making your selfies look better, these filters are supposed to make AI behave better. These filters are built on a cocktail of things we humans call ethical standards:Societal Norms: Basically, whats considered normal and acceptable in society. Think holding doors open for people, saying please and thank you, and not, you know, generating hate speech (JarvisLabs, 2024). These norms, however, are often skewed towards where the AI is developed kinda like how American TV shows dominate global streaming, right?Legal Standards: Laws are like the ultimate do not cross lines. AI models are supposed to steer clear of illegal stuff like defamation, inciting violence, or anything that lands you in actual jail (Solan, 2002). No AI jail just yet, thankfully.Ethical Guidelines: These are the shoulds and oughts of AI. Think fairness, transparency, accountability the kind of stuff Ive been researching and writing about in my Medium blogs. Its about making AI thats not just smart, but also well, decent (Floridi, 2023).Company Values: Big companies like Google AI and others have their own ethical principles. Its like their AIs personality they want their models to reflect their brand, their image, their vibe (Google AI, 2024).Trivia Time: The debate around AI censorship is older than you think! Sci-fi has been wrestling with this for decades, from 2001: A Space Odysseys HAL 9000 to Westworlds rogue robots!AI Filters: Making models behave, one filter at a time.The goal of all this alignment is simple: make AI a force for good, build trust, and avoid robot uprisings you know, the usual (Russell, 2019). Companies like OpenAI are pouring resources into this, trying to make ChatGPT and its buddies the responsible citizens of the AI world (OpenAI, 2024c).But heres the plot twist, folks. Censorship, even in AI, can backfire. Go too strict, and you get over-alignment (Carlini et al., 2020). Imagine a chatbot so worried about being offensive, it refuses to talk about anything remotely interesting or controversial. Hey AI, tell me about the French Revolution. Im sorry, but discussing revolutions might be triggering for some users. Can I interest you in cat videos instead? Frustrating, right? And kinda useless (Digit, 2025). Its like that friend whos so afraid of saying the wrong thing, they end up saying nothing at all.Plus, who decides whats ethical anyway? Ethics are like fashion trends they change, they vary across cultures, and whats cool in California might be cringe in Kolkata. Embedding specific ethics into AI can lead to bias, and suddenly, your aligned AI is just aligned to one viewpoint (Birhane et al., 2021). Not exactly the unbiased, all-knowing oracle we were promised.Pro Tip: Remember, AI is a tool. Like any tool, it can be used for good or bad. Its about responsible use, not just censorship.And heres a thought that keeps me up at night are we stifling innovation with all this censorship? Are we missing out on breakthroughs because were too scared of AI stepping out of line? Are we creating an innovation bottleneck in our rush to be responsible? (Vincent, 2023). Its like telling a rockstar to only play elevator music technically music, but kinda missing the point, right?Trivia Time: Did you know that early internet pioneers were fiercely against censorship? They envisioned the web as an uncensored space for free information flow. AI is kinda facing a similar crossroads now.The best way to predict the future is to create it. Peter Drucker.(But we gotta create it responsibly, folks! Dr.Sewak)Now, flip the coin. Uncensored AI. Sounds edgy, sounds rebellious, sounds like trouble? Maybe. But also, maybe opportunity? These models are designed to be information free-flow zones (Perez, 2024). Think of them as the internet, unfiltered, in AI form. Their core principles are:Information Freedom: Like the ACLU says, free speech is kinda a big deal. Uncensored AI leans into that, aiming for unrestricted access to info and ideas. Its like the AI version of a public library, with no librarian telling you what you cant read.Comprehensive Data Dive: These models want to use all the data theyre trained on the spicy memes, the controversial debates, the weird corners of the internet. No cherry-picking, no sanitizing. Its like saying, Give me the whole buffet, even the questionable-looking tuna casserole! (Luccioni et al., 2021).Exploration Powerhouse: Uncensored AI can dive into topics aligned models might shy away from. Think cutting-edge research, exploring taboo subjects, pushing creative boundaries. Its like letting a scientist loose in a lab with no safety goggles potentially dangerous, but maybe theyll discover something amazing (Vincent, 2023).User Control Nirvana: Imagine AI that you can actually control. Want it vanilla? Censored-lite? Full-on uncensored chaos? You get to choose! Its like having a volume knob for AI ethics crank it up or down as you please (Wolfram, 2023).AI Censorship: Dial it up, dial it down user control is the name of the game?Sounds like a utopia of information, right? Well not so fast. Uncensored AI is also a potential minefield. Remember that reckless friend? Yeah, they can also crash your car. The risks are real, and theyre kinda scary:Hate Speech Bonanza: Uncensored models can spew out racist, sexist, and hateful garbage faster than a Twitter argument. This aint just theoretical it can fuel online harassment, discrimination, and real-world harm (Gebru et al., 2018).Misinformation Mayhem: Fake news? Propaganda? Conspiracy theories? Uncensored AI is like a super-spreader event for all of the above. It can erode trust faster than a politicians promise and mess with everything from elections to public health (Pennycook & Rand, 2019).Weaponized AI: Cybercrime, deepfakes, sophisticated scams uncensored AI can be weaponized by bad actors faster than you can say cybersecurity breach (Brundage et al., 2018). Remember that CNN article I read? ChatGPT Is a Cyber Weapon Waiting to Be Weaponized (Metz, 2023). Chilling, right?Trust Gone with the Wind: If uncensored AI runs wild, spewing garbage and causing chaos, public trust in all AI even the good stuff will vanish faster than free pizza at a tech conference (Cave & Dihal, 2023). And thats bad news for everyone in the AI game.Open Source = Open Season? Many uncensored models are open source, which is awesome for access and innovation, but also kinda risky? Malicious folks can tinker with them, exploit vulnerabilities, and unleash AI mayhem without anyone really knowing who to blame (Dark Reading, 2025; Orca Security, 2024). And open-source licenses? Dont even get me started on the legal headaches (Contreras, 2015).Pro Tip: Open source AI is like open source software powerful and democratizing, but needs community responsibility and oversight.Trivia Time: The term open source comes from software development, but its now transforming fields from biology to AI!With freedom comes responsibility. Eleanor Roosevelt.(Uncensored AI needs responsibility baked in, not bolted on. Dr.Sewak)So, were stuck between a rock and a hard place, right? Censorship too much, innovation dies. Censorship too little, chaos reigns. Is there a middle path? Is there a way to get the best of both worlds the power of uncensored AI, minus the apocalypse? I think there is. And its called arbitration.Stay tuned, folks, because things are about to get arbitrated!Arbitration: A New Frontier in Governing Uncensored ModelsDelving into a structured framework to address the unique challenges of unfiltered artificial intelligence.Okay, so weve painted a picture of uncensored AI as this wild, untamed beast powerful, yes, but also potentially bitey. Weve seen that censorship alone isnt the answer. So, what is the answer? Drumroll please Arbitration! Yeah, I know, sounds kinda legal-y and boring. But trust me, in the AI world, arbitration is about to become the new black.Pro Tip: Arbitration isnt just for legal eagles. Its a problem-solving tool, a way to navigate complex issues fairly and effectively.Trivia Time: Arbitration is ancient! It dates back to ancient Greece and Rome, way before AI or even the internet was a twinkle in anyones eye. Humans have been squabbling and needing mediators for millennia!The arc of the moral universe is long, but it bends towards justice. Martin Luther King Jr.(Arbitration is about bending the AI universe towards justice and responsibility. Dr.Sewak)Now, forget those images of stuffy courtrooms and endless paperwork. In the context of uncensored AI, arbitration is something different. Its a structured, independent process for sorting out the messes that uncensored AI might create. Think of it as AIs ethical debugger? Yeah, lets go with that. Heres the breakdown of what AI arbitration, in my Dr. Sewak-approved version, actually means:Neutral Third Party to the Rescue: Imagine two AI models having a digital shouting match (it happens, trust me). Arbitration brings in a neutral referee someone whos not on either side, someone impartial to make a fair call. This is crucial because self-regulation in the AI wild west? Yeah, good luck with that. We need someone objective (IndiaAI, 2024; JDSupra, 2024).Expert Brainpower Unleashed: These arent just any referees. Were talking AI ethics gurus, tech law ninjas, cybersecurity commandos, and domain experts from every field you can imagine (Norton Rose Fulbright, 2024). Think of it as the Avengers of AI ethics, assembled to tackle the toughest AI dilemmas. Expertise is key because AI problems are well, complex.Flexibility is the Name of the Game: Arbitration isnt one-size-fits-all. Its like a Swiss Army knife of dispute resolution. Mediation for the chill disputes, expedited processes for the urgent ones, and full-blown hearings for the big kahunas. AI problems are diverse, so the solutions need to be too (PON, 2025; Norton Rose Fulbright, 2024). AI can even help manage these cases AI-powered case management? Mind blown! (IndiaAI, 2024; JDSupra, 2024).Fix the Problem, Prevent the Future: Arbitration isnt just about slapping wrists. Its about fixing the harm now and stopping it from happening again. Its about recommending fixes, setting guidelines, and nudging the whole AI ecosystem towards responsible behavior. Its like AI therapy, but for the whole industry.Transparency Sort Of: Okay, full transparency might be tricky (trade secrets, privacy, etc.). But the process should be transparent, and the outcomes as transparent as possible. Accountability is key, even for AI. We need to build trust, not just black boxes.The AI Arbitration Dream Team: Expertise, impartiality, and a whole lot of brainpower.So, how does this AI arbitration magic actually work? Lets break down the key ingredients of this framework, Dr. Sewak style:3.2.1. The Avengers Assemble: Multi-Stakeholder Arbitration PanelThink of this panel as the Justice League of AI ethics. We need a diverse crew, folks, not just tech bros in hoodies. This panel needs:AI Ethics Gurus: These are the folks who actually think about AI ethics all day. They know their Kant from their Keras, their utilitarianism from their well, you get the idea. Theyre the moral compass of this whole operation.Legal Eagles: Lawyers, legal scholars the folks who speak legalese fluently. Theyll navigate the legal minefields, figure out liability, and make sure our arbitration process is, you know, legal.Tech Wizards: We need the techies, the coders, the model whisperers. They understand the nuts and bolts of AI, the black boxes, the algorithms gone wild. Theyll be the translators between the ethical and legal folks and the actual AI tech.Domain Deep Divers: AI is everywhere, from healthcare to finance to cat video recommendations. We need experts from different fields to understand the context of AI harms. A healthcare AI gone wrong? Needs healthcare experts. A finance AI causing market mayhem? Bring in the finance gurus.Public Voice Amplifiers: Lets not forget the people! We need reps from civil society, consumer groups, the folks who actually use AI and get affected by it. Their voice matters, big time.Pro Tip: Diversity isnt just a buzzword. Its essential for fair and effective AI governance. Different perspectives, different expertise thats where the magic happens.Trivia Time: The concept of multi-stakeholder governance is hot in tech right now. Its about bringing everyone to the table industry, government, civil society, academia to solve complex problems together.None of us is as smart as all of us. Ken Blanchard.(Teamwork makes the AI dream work responsibly. Dr.Sewak)This panel needs to be squeaky clean independent, impartial, no conflicts of interest. Think of it as the AI version of Switzerland neutral and trustworthy. And term limits? Yes, please! Fresh blood, fresh perspectives, keep it dynamic.3.2.2. Complaint Central: Dispute Intake and AssessmentOkay, panels assembled. Now, how do people actually file a complaint when uncensored AI goes off the rails? We need a system, folks, not just frantic emails and angry tweets. Think Complaint Central, AI edition:Multiple Entry Points: Online forms, email, maybe even a Bat-Signal for really urgent cases. Make it easy for everyone to report issues users, developers, random folks on the internet who spot something fishy.Standardized Forms No Mystery Novels: Complaint forms should be clear, simple, and capture the key info what happened, which AI model, who got hurt, evidence, the whole shebang. No cryptic riddles, please.Triage Time Sorting the Signal from the Noise: Not every complaint will be legit. We need a quick filter to weed out the my AI insulted my cat complaints from the this AI is spreading election misinformation emergencies. AI can even help here AI triage? Meta-AI! But human eyes on the prize, always.Investigation Mode Sherlock Holmes vs. AI: Serious complaints? Time for a deep dive. Evidence gathering, interviews, tech forensics the full detective toolkit. The arbitration panel oversees this, making sure its fair, thorough, and not a witch hunt.Pro Tip: Make reporting easy and accessible. The more eyes and ears on the ground, the better we can catch AI issues early.Trivia Time: Triage comes from battlefield medicine! Its about prioritizing cases based on urgency. Makes sense for AI harms too, right?Justice delayed is justice denied. William E. Gladstone.(We need a system thats not just fair, but also efficient. Dr.Sewak)3.2.3. Flexibility is the Superpower: Arbitration ProceduresAI problems are like snowflakes no two are exactly alike. So, arbitration needs to be flexible. Think choose your own adventure dispute resolution:Mediation Magic Lets Talk it Out: Sometimes, all it takes is a neutral mediator to get people talking, to find common ground. Mediation is like AI couples therapy helping parties understand each other, find solutions, and maybe even hug it out at the end (PON, 2025). AI can even assist in mediation analyzing communication, suggesting compromises (IndiaAI, 2024)! Mind. Blown. Again.Expedited Express Lane Speedy Justice: Small disputes, urgent issues? Expedited arbitration is your friend. Fast track procedures, streamlined evidence, quicker decisions. Think AI dispute resolution on fast-forward.Full-Blown Arbitration Arena When Things Get Serious: Big, complex, contentious cases? Bring out the full arbitration machinery. Formal hearings, legal arguments, expert testimony, the whole nine yards. Think AI court, but hopefully less dramatic than TV court.Advisory Wisdom Guidance, Not Orders: Sometimes, we just need advice. For novel ethical dilemmas, for cutting-edge tech issues, the panel can issue non-binding advisory opinions. Think AI Yoda, dispensing wisdom to guide the AI galaxy.Pro Tip: Flexibility is key. Match the dispute resolution process to the type of dispute. Not every AI spat needs a full-blown legal battle.Trivia Time: Mediation and arbitration are often used interchangeably, but theyre different! Mediation is about facilitation, arbitration is about adjudication. Now you know!The greatest remedy for anger is delay. Seneca.(But in the AI world, sometimes speed is also of the essence. Balance, people, balance! Dr.Sewak)The panel gets to decide which procedure fits best for each case. Theyre the AI dispute resolution chefs, choosing the right recipe for each problem.3.2.4. Fix it and Forget it? Nope! Remediation and GovernanceArbitration isnt just about slapping wrists and saying dont do it again. Its about fixing the problem and preventing future screw-ups. Think remediation and AI governance recommendations! Catchy, right? Heres what it means:Remediation Rumble Making Things Right (Now): Harm done? Arbitration panel to the rescue! They can recommend (or even mandate!) fixes: Model Makeovers: AI model, youre grounded! Go back to training and fix your biases! Panel can order developers to tweak the model, refine the data, add safety features without going full censorship. Usage Limits Time Out for AI: AI model, youre banned from Twitter for a week! Usage restrictions, especially in high-risk areas. Think AI probation. Show Me the Money Compensation for Harm: Real harm, real consequences. Panel can recommend compensation for victims. AI accountability with teeth. Public Apology Tour Say Youre Sorry, AI: AI model, issue a public apology for spreading misinformation! Public apologies, corrections transparency and making amends.Pro Tip: Remediation is about fixing the immediate problem. Governance recommendations are about preventing future problems. Both are crucial for responsible AI.Governance Guidance Setting the Rules of the AI Road (for the Future): Arbitration isnt just reactive. Its proactive too! The panel becomes the AI governance guru, dishing out wisdom to: Developers: Hey AI devs, heres a playbook for responsible uncensored AI development! Data curation tips, safety testing checklists, transparency guidelines the whole shebang! Deployers: Companies using uncensored AI, listen up! Risk assessment frameworks, user agreements that actually make sense, content moderation strategies that dont kill innovation get on board! Policymakers: Governments, regulators, are you listening? Heres what weve learned, heres what works, heres how to regulate AI without stifling progress! Policy recommendations, AI style! Standard Setters: Industry standards, anyone? Lets create some benchmarks for responsible uncensored AI! Voluntary standards, industry-wide guidelines lets raise the bar!Trivia Time: Governance sounds boring, but its actually about shaping the future! Think of it as designing the rules of the AI game, making sure its a game we all want to play.An ounce of prevention is worth a pound of cure. Benjamin Franklin.(Governance recommendations are the ounce of prevention for uncensored AI. Dr.Sewak)Arbitration isnt just about solving todays AI headaches. Its about building a better AI future, one arbitration case at a time.3.2.5. Shine a Light: Transparency and AccountabilitySunlight is the best disinfectant, right? Transparency and accountability are the secret sauce of effective arbitration. We need to build trust, and trust comes from well, trust and transparency (and maybe a little bit of magic).Process Transparency Open the Black Box (a Little): Arbitration procedures, rules, panel member selection all public knowledge. No secret AI cabals here. Fairness needs to be seen to be believed.Outcome Transparency Share the Wisdom (Carefully): Case summaries, anonymized decisions, key findings lets share the lessons learned. But confidentiality matters too. Trade secrets, privacy gotta balance openness with discretion.Accountability Anchors Making it Stick: Arbitration decisions are great if they actually mean something. We need ways to ensure compliance:Voluntary Buy-In Good AI Citizens: Industry norms, ethical peer pressure, reputational points lets encourage voluntary compliance. Make responsible AI the cool kid club.Contractual Clout Ink it and Link it: Arbitration clauses in user agreements, developer contracts make arbitration legally binding, at least in some contexts.Regulatory Backup Government to the Rescue? Maybe, maybe not. Regulators could recognize arbitration decisions, give them some legal teeth. Tricky balance though dont want to stifle innovation with red tape.Public Shaming (and Praising) Community Power: Public reports, monitoring, community watchdogs sunlight as disinfectant, remember? Shine a light on good AI actors and less good ones.Pro Tip: Transparency and accountability build trust. Trust is essential for the long-term success of AI.Trivia Time: Transparency is a buzzword, but its also a fundamental principle of good governance, from democracies to AI arbitration frameworks!Trust, but verify. Ronald Reagan (and a Russian proverb).(Transparency is the verify part of AI trust. Dr.Sewak)Arbitration needs to be more than just talk. It needs to be real, effective, and accountable. Transparency is how we get there.And guess what? AI can even help run arbitration! Meta-meta-AI!3.3. AI-Powered Arbitration: Robot Referees?Yeah, you heard that right. AI can actually make arbitration better? More efficient? Less human-error-prone? Maybe. Think AI as the arbitration sidekick:Case Management Bots AI Paper Pushers: Automate the boring stuff case intake, document wrangling, scheduling, reminders. Let AI handle the paperwork, humans handle the human stuff (JDSupra, 2024).Evidence Analysis AI Data Detective: Mountains of text, audio, video evidence? AI can sift through it, find patterns, summarize key points, flag inconsistencies. Think AI Sherlock Holmes, but for data (IndiaAI, 2024). NLP and machine learning to the rescue!Legal Research AI Lawyer in a Box: Arbitrators need to know the law, precedents, ethical guidelines. AI legal research tools can speed things up, find relevant cases faster than you can say amicus curiae.Mediation AI-ssistant Peace-Making Algorithms: AI can analyze communication in mediation, spot roadblocks, suggest compromises, nudge parties towards agreement (PON, 2025). Think AI relationship counselor, but for AI disputes.Bias-Busting AI Fairness Enforcer: Irony alert! AI can help detect and mitigate bias in the arbitration process itself (Norton Rose Fulbright, 2024). Bias in arbitrator selection? AI can flag it. Bias in evidence interpretation? AI can help spot it. But careful here. AI bias-detectors need to be bias-free themselves! Its bias-ception!Pro Tip: AI can augment arbitration, but it shouldnt replace humans. Human judgment, ethical reasoning, empathy still essential.Trivia Time: AI is already being used in real-world arbitration and mediation (e.g. IndiaAI ADR Mechanism, JAMS Resolution Center). The future is now, folks!Technology is best when it brings people together. Matt Mullenweg.(AI can bring people together to resolve AI disputes! Dr.Sewak)Irony, again!AI in arbitration its not about robot judges taking over. Its about humans and AI working together to make the process fairer, faster, and more effective. Its about harnessing AI to govern AI. Full circle, baby!Harnessing Power, Mitigating Risks: Real-World Arbitration in ActionExploring the potential benefits and perils of unrestricted AI and how arbitration can bridge the gap.Theory is cool, but real-world examples? Thats where the rubber meets the road. Lets dive into some case studies, Dr. Sewak style, to see how this arbitration thing might actually work in practice. Think of these as AI dispute hypotheticals, inspired by real-world AI scenarios and the kind of stuff Ive seen brewing in the AI world.Pro Tip: Case studies arent just stories. Theyre thought experiments, ways to test ideas and anticipate future challenges.Trivia Time: Case studies are a staple in business schools and law schools. Now, theyre becoming essential for AI ethics and governance too!Tell me and I forget. Teach me and I remember. Involve me and I learn. Benjamin Franklin.(Case studies are about involving us in the learning process. Dr.Sewak)4.1. Case Study 1: Open-Source LLMs Gone Rogue Misinformation MayhemThe Setup: Imagine a top AI lab releases an open-source, uncensored Large Language Model (LLM). Think Metas Llama 2, but even less filtered. The goal? Innovation, open access, AI for the people! Sounds noble, right? Well this model becomes a misinformation factory overnight. Bad actors weaponize it to pump out hyper-realistic fake news during a crucial election. Social media explodes with AI-generated propaganda, trust in institutions crumbles, democracy well, you get the picture (Dark Reading, 2025).Misinformation Machine: When open-source AI becomes a propaganda factory.The Harm: Massive misinformation, manipulated public opinion, democratic meltdown, maybe even real-world violence. Not good.Arbitration to the Rescue? Enter the AI Arbitration Panel! Who files a complaint? Affected citizens, civil society groups, maybe even governments. The panel investigates: Did this open-source model directly cause the misinformation storm? Whose responsibility is it? The AI lab that released it? The malicious actors who weaponized it? Both?Remediation Rumble:Tech Fixes (Sort Of): Cant really censor open-source code once its out there. But maybe the panel recommends open-source safety patches, AI tools to detect AI-generated misinformation. Think open-source antibodies for the AI misinformation virus.Governance Guidance Open Source Rules of the Road: Panel issues recommendations to AI labs, open-source communities: Responsible release practices for uncensored models, folks! Risk assessments, transparency, community moderation make it happen!Public Awareness Blitz: Panel pushes for public education campaigns: Hey citizens, AI misinformation is real! Learn to spot it, think critically, be media literate!Relevant Companies/Models: Metas Llama 2, Hugging Face open-source models, the whole open-source LLM ecosystem.Dr. Sewak Says: Open source is awesome, but open responsibility needs to be part of the deal. Think open source software security its a community effort. Same for AI ethics.4.2. Case Study 2: Creative AI Copyright Clash Artistic Anarchy?The Setup: Startup X launches an uncensored AI model for creative content writing, music, visual art, the works. Its a hit! Users are creating amazing, novel stuff. But wait a minute. Some of this original AI art looks suspiciously familiar. Turns out, the AI is kinda borrowing a lot from copyrighted material. Content creators, copyright holders theyre not happy (Stability AIs Stable Diffusion is kinda in this space, just sayin).AI Artist or AI Art Thief?The Harm: Copyright violations, artists losing income, creative industries in chaos, legal battles galore. Not a pretty picture.Arbitration to the Rescue? Copyright holders, artists file complaints. Arbitration panel copyright law experts, AI techies investigates: Is this real copyright infringement? Is AI transformative use or just plain copying? Wheres the line?Remediation Rumble:Copyright Court, AI Edition: Panel assesses copyright infringement, applies fair use doctrines, figures out if AI crossed the line.Licensing Deals AI Art Royalties? Panel recommends licensing agreements between AI startup and copyright holders. AI art royalties new revenue stream for artists? Maybe.Governance Guidance Copyright Rules for AI Creators: Panel issues guidelines for AI developers: Copyright compliance checklist for creative AI! Ethical data sourcing, infringement detection mechanisms get it done! Maybe even ethical AI dataset standards (Solaiman et al., 2023).Relevant Companies/Models: Stability AIs Stable Diffusion, AI music generators, AI writing tools the whole creative AI boom.Dr. Sewak Says: AI creativity is cool, but copyright is still a thing. We need to figure out how AI and copyright can coexist peacefully. Think AI music licensing, AI art royalties new business models for a new era.4.3. Case Study 3: Uncensored AI Bias in High-Stakes Decisions Algorithmic DiscriminationThe Setup: Company Z deploys an uncensored AI model for loan applications, hiring, criminal risk assessment you know, the important stuff. Turns out, the uncensored model is biased. Surprise! It starts discriminating against certain demographic groups unfair loan denials, biased hiring, skewed risk scores. People are getting hurt, opportunities are being denied, and lawsuits are brewing (think COMPAS risk assessment tool, AI hiring platforms real-world examples of algorithmic bias).Algorithmic Justice? When uncensored AI decisions become discriminatory.The Harm: Systemic discrimination, inequality amplified, lives impacted, trust in AI in critical domains shattered.Arbitration to the Rescue? Affected individuals, civil rights groups file complaints. Arbitration panel algorithmic fairness experts, bias detection gurus, human rights lawyers investigates: Is this AI really biased? How bad is it? Whos responsible for fixing it?Remediation Rumble:Bias Busting Bootcamp for AI: AI model, report to debiasing bootcamp, stat! Panel orders model retraining with debiased data, fairness-aware algorithms (Mehrabi et al., 2021). AI bias correction, in real-time.Algorithmic Audits Check Under the AI Hood: Panel mandates ongoing algorithmic audits, bias monitoring. Company Z, show us your AI fairness report card! Regularly!Human Override Button Humans in the Loop: In high-stakes decisions, humans must have the final say! AI is a tool, not a dictator! Human oversight, human accountability.Compensation for Victims Making Amends for AI Bias: Panel recommends compensation for those harmed by AI discrimination. Putting a price on AI-driven unfairness.Governance Guidance Fairness First in AI Deployment: Panel issues guidelines for organizations deploying AI in high-stakes areas: Ethical AI frameworks are not optional! Fairness, transparency, accountability bake it in from day one! Promote ethical AI standards (Mkander & Hagras, 2022).Relevant Companies/Models: AI in finance, HR tech, criminal justice domains where bias is a critical concern.Dr. Sewak Says: Algorithmic bias is real. And uncensored AI can amplify it big time. Fairness needs to be baked into AI from the start, not bolted on as an afterthought. Human oversight is non-negotiable in high-stakes AI decisions.These case studies, while hypothetical, are rooted in real-world AI challenges. They show that uncensored AI, while powerful, needs governance. And arbitration, my friends, might just be the new frontier in making that governance a reality.Challenges and the Path Forward: Navigating the AI Governance MazeAcknowledging limitations and discussing future directions for arbitration in the age of uncensored AI.Lets be real, folks. This arbitration for uncensored AI thing? Its not a magic wand. Its not a perfect solution. Its a framework, a proposal, a work in progress. There are challenges, speed bumps, and maybe even a few AI-sized potholes on this road to responsible AI governance. Lets face them head-on, Dr. Sewak style.Pro Tip: No solution is perfect. The goal is to make things better, not perfect. Progress, not perfection.Trivia Time: Challenges are just opportunities in disguise, right? At least, thats what motivational posters tell us. In AI, challenges are definitely opportunities for innovation and problem-solving.The impediment to action advances action. What stands in the way becomes the way. Marcus Aurelius.(AI governance challenges? Lets turn them into our way forward. Dr.Sewak)5.1. Defining Unacceptable Harm: The Fuzzy LineWhat is unacceptable harm from AI, anyway? Its not like breaking a law or causing physical injury. AI harms can be fuzzy. Subjective. Context-dependent (Vallor, 2016). Whats harmful to you might be free speech to someone else. And AI operates in a million different contexts. Defining harm in AI is like nailing jelly to a wall.The Fuzzy Line: Defining unacceptable harm in the age of uncensored AI.Dr. Sewaks Mitigation Strategies:Harm-O-Meter 3000 (Multi-Dimensional Edition): We need a way to measure harm thats more nuanced than just yes/no. Think: Severity Scale: Minor annoyance to societal collapse harm isnt binary. Probability Factor: Likelihood of harm actually happening. Potential harm vs. actual harm. Vulnerability Index: Whos most likely to get hurt? Kids? Marginalized groups? Factor in vulnerability. Context Compass: Context, context, context! Same AI output, different contexts, different harm levels. Intent-O-Meter: Was harm intended? Or just AI being AI? Intent matters, sometimes.Ethical Barometer Community Standards Check: Tap into ethical guidelines, human rights frameworks, community norms. Crowdsource wisdom, dont just dictate from on high. Inclusivity is key.Case-by-Case Wisdom Human Judgment Still Matters: No algorithm can define harm perfectly. Arbitration panel needs human judgment, ethical reasoning, common sense. AI can assist, not replace, human wisdom.5.2. Enforcement Enigma The Decentralized AI Wild WestUncensored AI is often open source, decentralized, global. Enforcing arbitration decisions in this wild west? Good luck, right? Its not like we can send AI police to arrest a rogue algorithm. Traditional enforcement tools kinda useless here.Enforcement Enigma: Policing the decentralized frontier of uncensored AI.Dr. Sewaks Mitigation Strategies:Soft Power Playbook Norms, Not Laws (Initially): Focus on soft law industry best practices, ethical peer pressure, reputational incentives. Make responsible AI the norm, the cool thing to do. Arbitration framework as a norm-setter.Contractual Chains Binding Agreements (Where Possible): Incorporate arbitration clauses into user agreements, developer contracts. Make arbitration decisions contractually enforceable, at least. Legal glue for the AI wild west.Tech Transparency Tools AI Audit Trails: Tech to the rescue! AI model watermarking, provenance tracking, blockchain registries for AI. Make AI more transparent, auditable, traceable. Tech-powered accountability.Global Governance Gang International AI Cooperation: AI is global, governance needs to be too. International cooperation, harmonized frameworks, cross-border arbitration. AI governance without borders.Community Cops Open Source Responsibility: Leverage the open-source AI community itself! Community moderation, code reviews, reputational systems. Open source AI, open source responsibility. Arbitration as a community tool.5.3. Bias in Arbitration Fairness ParadoxCan arbitration itself be biased? Irony alert! Arbitrators are human, AI tools can be biased (Norton Rose Fulbright, 2024). How do we ensure fairness in the arbitration process itself? Bias-ception, again!Bias-ception: Can AI arbitration be truly unbiased? The fairness paradox.Dr. Sewaks Mitigation Strategies:Diversity Deluxe Panel Represent All Voices: Diverse arbitrators expertise, background, gender, race, geography. Diversity as a bias-buster.Transparent Arbitrator Selection Open and Honest Process: Clear criteria, conflict-of-interest checks, stakeholder input. No backroom deals, no secret AI cabals.Bias-Proof AI Tools Audit the Auditors: If AI helps arbitration, audit those AI tools for bias! Bias detection for bias detectors. Meta-bias-busting!Human in the Loop Judgment, Not Just Algorithms: Human oversight, human review, human judgment. AI assists, humans decide. Human wisdom still essential.Appeals Process Second Opinions Welcome: Appeals mechanisms, review boards. Checks and balances for arbitration itself. Accountability for the arbitrators too.5.4. Tech Tsunami Keeping Up with AI EvolutionAI is evolving faster than my caffeine addiction. Arbitration framework needs to keep pace. Outdated framework = useless framework. Adaptability is not optional, its survival.Future-Proof Arbitration: Adapting to the ever-evolving AI tsunami.Dr. Sewaks Mitigation Strategies:Regular Reboot Cycles Framework Updates, Annually (or Faster): Annual reviews, updates, revisions. Framework needs to be a living document, not a museum piece. Agile AI governance.Modular Design Plug-and-Play Governance: Modular framework architecture. Add new procedures, expertise areas, recommendations as needed. Flexibility built-in.Continuous Learning Loop AI Governance Research Lab: Research, monitor AI trends, analyze arbitration cases, learn from experience. AI governance as a continuous learning process. AI governance R&D.Pilot Programs Test in the Wild: Pilot arbitration frameworks in real-world settings. Test, refine, iterate. Real-world feedback loop.AI Research Collabs Partner with the Brains: Collaborate with AI researchers, ethicists, legal scholars. Stay ahead of the curve, tap into the best minds. AI research-governance synergy.Pro Tip: AI governance needs to be as dynamic and adaptive as AI itself. Static frameworks are doomed to fail.Trivia Time: Agile and iterative are buzzwords from software development. But theyre also key to effective AI governance in a fast-paced world.It is not the strongest of the species that survives, nor the most intelligent that survives. It is the one that is most adaptable to change. Charles Darwin.(Adaptability is survival in the AI governance jungle. Dr.Sewak)Arbitration for uncensored AI? Its not a perfect solution, but its a necessary solution. Its a work in progress, a journey, a new frontier. And like any frontier, its gonna be challenging, messy, and maybe a little bit wild west-y. But with the right framework, the right experts, and a whole lot of responsible intention, we can harness the power of uncensored AI while keeping the chaos at bay. Its a tightrope walk, folks, but its a walk worth taking. For the future of AI, for the future of well, everything. Lets arbitrate, shall we?Conclusion: Towards a Responsible Future for Uncensored AISummarizing the key takeaways and reiterating the importance of arbitration for responsible AI governance.Alright, folks, weve reached the summit of our AI arbitration mountain. Weve explored the wild landscape of uncensored AI, faced the challenges head-on, and mapped out a potential path forward arbitration. Lets zoom out for a birds-eye view, Dr. Sewak style, and recap the key takeaways.The Summit of Responsible AI: Arbitration as our path to a balanced future.Dr. Sewaks Key Takeaways The Arbitration Advantage:Balance is the Name of the Game: Uncensored AI is a double-edged sword. Arbitration offers a way to wield that sword responsibly, balancing innovation with safety, freedom with ethics. Its about finding that sweet spot, not swinging wildly in either direction.Expertise Matters, Impartiality is Essential: AI problems are complex. Arbitration brings in the Avengers of AI ethics diverse experts, neutral perspectives, informed decisions. No more AI governance by gut feeling.Flexibility Wins the Day: AI disputes are diverse. Arbitration is adaptable mediation, expedited processes, full hearings, advisory opinions. One-size-fits-none in AI governance.Fix Today, Prevent Tomorrow: Arbitration isnt just damage control. Its about remediation and governance. Fixing harms and shaping responsible AI practices for the future. Proactive, not just reactive.Transparency Builds Trust, Accountability Makes it Real: Transparency in process, accountability in outcomes. Arbitration aims for both. Trust and accountability the twin pillars of responsible AI.The Road Ahead Challenges and Opportunities:Defining Harm Still Fuzzy, Still Crucial: Defining unacceptable harm remains a challenge. Multi-dimensional assessment, community standards, human judgment the path forward.Enforcement in the Wild West Decentralization Dilemma: Enforcing arbitration in the decentralized AI landscape is tricky. Soft law, contracts, tech transparency, global cooperation, community responsibility the multi-pronged approach.Bias in Arbitration Fairness Paradox Persists: Bias can creep into arbitration itself. Diversity, transparency, AI audits, human oversight, appeals bias mitigation strategies are essential.Keeping Up with the AI Tsunami Adaptability Imperative: AI evolves fast. Arbitration frameworks must evolve faster. Regular updates, modular design, continuous learning, pilot programs, research collaborations adaptability is survival.The Dr. Sewak Vision Responsible AI Future:Arbitration for uncensored AI isnt a perfect solution, but its a vital step. Its a move towards a more responsible, more ethical, more governed AI future. Its about harnessing the immense power of AI while safeguarding human values, promoting innovation responsibly, and ensuring that AI serves humanity, not the other way around. Its a tightrope walk, yes. But its a walk we must take. For the future of AI, and for the future of us all. Lets build this new frontier together, responsibly, ethically, and with a healthy dose of arbitration, just in case things get a little uncensored. What do you say, friend? Ready to arbitrate the future?Pro Tip: The future of AI governance is not about finding a perfect solution, but about building resilient, adaptable, and ethically informed frameworks.Trivia Time: The frontier metaphor is apt for AI. Like the Wild West, AI is a new territory, full of opportunity and risk. We need to build our AI towns responsibly, with rules, sheriffs (arbitrators?), and a sense of community.The best is yet to come. Frank Sinatra.(Lets make sure the best AI future is also a responsible AI future. Arbitration can help us get there. Dr.Sewak)And thats a wrap, folks! Dr. Sewak, signing off. Go forth and arbitrate responsibly! And maybe, just maybe, we can tame this uncensored AI beast and build a future where AI is both powerful and good. Until next time, keep those algorithms ethical, and those punchlines punchy!7. References7.1. AI Ethics and GovernanceFloridi, L. (2023). The ethics of artificial intelligence. Oxford University Press.Russell, S. J. (2019). Human compatible: Artificial intelligence and the problem of control. Penguin.Pfeiffer, J., Gutschow, J., Haas, C., Mslein, F., Maspfuhl, O., Borgers, F., & Alpsancar, S. (2023). Algorithmic fairness in AI: an interdisciplinary view. Business & Information Systems Engineering, 65(2), 209222.Google AI. (2024). AI at Google: Our Principles.OpenAI. (2024a). Our approach to AI safety.OpenAI. (2024b). Safety and Policy.7.2. Uncensored AI Models and RisksJarvisLabs. (2024, December 3). Uncensored LLM Models: A Complete Guide to Unfiltered AI Language Models.Zem, O. (2024, February 29). Uncensored Models in AI. Medium.Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., & Liang, P. (2021). On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258.Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P. S., & Gabriel, I. (2021). Ethical and social risks of harm from language models. arXiv preprint arXiv:2112.04359.Lyngass, S. (2023, March 16). ChatGPT is the new crypto: Meta warns hackers are exploiting interest in the AI chatbot. CNN Business.Disclaimers and DisclosuresThis article combines the theoretical insights of leading researchers with practical examples, and offers my opinionated exploration of AIs ethical dilemmas, and may not represent the views or claims of my present or past organizations and their products or my other associations.Use of AI Assistance: In the preparation for this article, AI assistance has been used for generating/ refining the images, and for styling/ linguistic enhancements of parts of content.License: This work is licensed under a CC BY-NC-ND 4.0 license.Attribution Example: This content is based on [Title of Article/ Blog/ Post] by Dr. Mohit Sewak, [Link to Article/ Blog/ Post], licensed under CC BY-NC-ND 4.0.Follow me on: | Medium | LinkedIn | SubStack | X | YouTube |Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

If you're in the market for a new Switch console, here's a deal worth consideration. AliExpress is currently offering a Nintendo Switch OLED console for only $223.61 after coupon code: USAFF30 is applied during checkout. This product is also stocked in a US warehouse, so it ships free and arrives within about a week. AliExpress has a 15-day free return policy and guarantees a refund if your order doesn't arrive within 20 days.This is a brand new, genuine Nintendo Switch console, but keep in mind that it is imported from either Japan or Hong Kong. What that means is that the warranty it comes with will most not likely be valid in the United States. The packaging and materials (like the manual) might also be written in another language). Functionally, however, Switch consoles are region unlocked so you'll be able to play them in the US without any problems. You'll be able to select English as your default language just like any console you buy here. A US power plug adapter is also supplied.Nintendo Switch OLED for $223.61Nintendo Switch OLED (Import)If you can afford the small price premium of the Switch OLED over the non-OLED model, then I think it's worth the upgrade. The Nintendo Switch OLED features an OLED display that offers better image quality than the original LCD display. That includes deeper blacks, more vibrant colors, and a better color gradient. The text is sharper and easier to read as well. The screen isn't that much bigger (7" on the OLED vs 6.2" on the non-OLED), but it does feel a lot bigger thanks to a thinner bezel. Other upgrades worth mentioning are a new and improved kickstand, a hard-wired ethernet port on the docking station, and an exclusive white Switch shell and matching Joy-Cons.$39.99 Nintendo Switch Video GamesSwitchPikmin 4SwitchSplatoon 3SwitchThe Legend of Zelda: Breath of the WildSwitchMario Kart 8 DeluxeSwitchKirby and the Forgotten LandBest Buy has dropped several must-have Nintendo Switch video games from $59.99 to $39.99, a savings of 33%. Several of them are available in your choice of a physical copy or digital download. You won't lose out even if you plan to upgrade to the Switch 2, since the new console has been confirmed to be backwards compatible with nearly all Switch games.The "Switch 2" Isn't Out Until June or LaterIf you're holding out on the next generation Switch 2 console, be prepared to wait for a while longer. It's possible that the console update might not be released until after June of 2025. We suspect the price of the Switch will be at least $400, so you'd be paying hundreds more than a Switch OLED console right now. Basically, if you don't own a Switch, now is still a good time to join the family. The games are still as amazing as when they were first released, and they're on sale more often than ever before. If you're looking for deals on accessories, check out the best Switch deals of 2024.Why Should You Trust IGN's Deals Team?IGN's deals team has a combined 30+ years of experience finding the best discounts in gaming, tech, and just about every other category. We don't try to trick our readers into buying things they don't need at prices that aren't worth buying something at. Our ultimate goal is to surface the best possible deals from brands we trust and our editorial team has personal experience with. You can check out our deals standards here for more information on our process, or keep up with the latest deals we find on IGN's Deals account on Twitter.Eric Song is the IGN commerce manager in charge of finding the best gaming and tech deals every day. When Eric isn't hunting for deals for other people at work, he's hunting for deals for himself during his free time.

When youre setting out to build or upgrade your gaming PC, the best graphics cards are often the first thing that comes to mind. Theres a simple reason for that: When it comes to PC games, GPUs are the most impactful component in determining your rig's raw frame rates. Quite simply, most of the time, a better graphics card directly results in better performance at least up to a point. With Nvidia's RTX 5090 and RTX 5080 cards out now (though it's nearly impossible to actually get one), here are the best graphics cards on the market. TL;DR: These Are the Best Graphics Cards:Our Top PickZotac Gaming Nvidia GeForce RTX 4070 SuperSee it at AmazonGigabyte Nvidia GeForce RTX 5090See it at NeweggGigabyte AMD Radeon RX 7900 XTXSee it at AmazonGigabyte AMD Radeon RX 7700 XTSee it at AmazonMSI Nvidia GeForce RTX 4060See it at AmazonThese days, GPUs have legitimately become a luxury good. With graphics cards like the Nvidia GeForce RTX 5090 costing upwards of $1,999, you can expect to pay top dollar for top performance. Prices are still way higher than they were when the GTX 970 blew my mind in 2014 even when adjusted for inflation. However, if you temper your expectations, you can still get a solid gaming experience for a fraction of that price, especially if you're okay with gaming at 1440p or 1080p. Ive been reviewing graphics cards for the last four generations, and Ive personally benchmarked, built with, and played games using every GPU on this list. However, if none of these strike your fancy, feel free to comment below what kind of gaming experience youre looking for, and Id be happy to help you find the perfect card for your build. What to Look for in a Graphics CardWhile it would be easy to just tell you to get the most powerful graphics card on the market for the best gaming experience, the truth is that picking a GPU is something you need to put a bit more thought into. not all graphics cards are created equal, you see, and everyone is going to need something a little different out of their PCI-E brick. The resolution you play games at is going to be the first thing you want to figure out. Whether you already have a gaming monitor you love or you're looking to build up an entire gaming battlestation, decide on your resolution first. This is because a graphics card that's great at 4K isn't exactly going to translate to an amazing 1080p graphics card. Just take a look at the Nvidia RTX 5090. While the new flagship flies at 4K, it can actually be slower than much cheaper graphics cards at 1080p, due to CPU bottlenecking. Instead, if you're playing at 1080p, something like the Intel Arc B580 is going to be a much better fit. You're still going to get solid gaming performance, but at a much lower price, which you can use to, well, buy more games. Likewise, 1440p gamers are probably best off buying something like the AMD Radeon RX 7700 XT or the Nvidia GeForce RTX 4070 Super. Budget is also a huge concern, and graphics cards are only getting more expensive these days. It would be awesome if everyone could afford an Nvidia GeForce RTX 5090, but that's just not the world we live in. These days, the floor sits around $200-$250. For that price you can get a solid 1080p graphics card, without having to go back to a previous generation. If you have a bit more cash, something like the Nvidia RTX 4060 will unlock all of Nvidia's exclusive bells and whistles though they're much less important on a lower-end GPU. If you really want to go all-out, you can get an amazing graphics card for around $1,000. Both the AMD Radeon RX 7900 XTX and Nvidia GeForce RTX 5080 will provide an awesome 4K gaming experience, and the one you choose should ultimately depend on how much you care about ray tracing. For most people that just want raw gaming performance, the Radeon RX 7900 XTX is probably going to be a better option, but you're going to want to wait a minute for AMD's next-generation offering. I don't know how well the Radeon RX 9070 XT is going to perform until I benchmark it, but it should prove to be great for 4K gaming, especially with the upcoming FSR 4.Luckily, it looks like this generation of graphics cards is going to make 4K gaming much more accessible. When I reviewed the Nvidia RTX 5070 Ti, I found that it had no problem breaking 60 fps in even demanding games like Black Myth Wukong. However, more expensive and powerful cards like the RTX 5080 and 5090 are going to give you more breathing room for future PC games. With more expensive graphics cards, however, power is going to be a big concern. You're going to want to make sure you check which power supply you have, and check it against the power requirements for the graphics card you're looking at. Something like the Intel Arc B580 can get away with a 450W PSU, but you're going to want to make sure you have something much more powerful for the Radeon RX 7800 XT, for instance. You don't need to go overboard and fork over the cash for a power supply that offers twice the recommended power, just make sure you have enough juice to keep your GPU going. Nvidia GeForce RTX 4070 Super Unboxing1. Nvidia GeForce RTX 4070 SuperThe Best Graphics Card for Most PeopleOur Top PickZotac Gaming Nvidia GeForce RTX 4070 SuperThis RTX 4070 Super from Zotac is the best graphics card for most people, packed with an awesome dual-fan cooler and all the performance you'll need.See it at AmazonProduct SpecificationsCUDA Cores/Stream Processors7168Base Clock1,980MHzBoost Clock2,475MHzVideo Memory12GB GDDR6XMemory Bandwidth504.2GB/sMemory Bus192-bitPower Connectors1 x 16-pinOutputs1 x HDMI 2.1, 3 x DisplayPort 1.4 Size10.5 x 4.4 x 1.7 inches (L x W x H)PROSGood power for the moneyCan break into 4K in some gamesCONSShould have had 16GB of VRAMWhile it kind of hurts to call any GPU that costs $599 affordable, thats kind of what the Nvidia GeForce RTX 4070 Super is going for. For a mid-range price, youre getting an extremely solid graphics card that can pump out 1440p-ready gaming performance, and can even stretch into 4K in certain games. Its not the most powerful graphics card out there right now, but this would be Goldilocks choice. When I reviewed the Nvidia GeForce RTX 4070 Super, I found it to be an extremely strong contender, and probably the best of the mid-generation refresh cards that Nvidia launched at CES 2024. While it still has the same 12GB of VRAM that held the original RTX 4070 back, the RTX 4070 Super is packed with 7,168 CUDA cores, compared to 5,888 in the original model. Thats a 21% jump in shading cores, and this difference is reflected in gaming performance.Nvidia RTX 4070 Super BenchmarksJust looking at a game like Cyberpunk 2077, which absolutely loves CUDA, you can see a 12% jump in performance over the RTX 4070 at 1440p. And, at 4K, the RTX 4070 Super is 13% faster than its predecessor in CD Projekt Reds demanding RPG. Likewise, when running a game like Forza Horizon 5 at 4K, the RTX 4070 Super manages an impressive 123 fps, compared to the 94 fps of the original RTX 4070. Thats a 30% increase in performance at the same launch price. No matter which game you play at 1440p, youre going to get extremely good performance out of the RTX 4070 Super, and thatll translate into 4K gaming more often than not. And with 1440p gaming growing in popularity, theres never been a better time to upgrade to a card that can really soar at that resolution. Nvidia GeForce RTX 5070 Ti Photos2. Nvidia GeForce RTX 5070 TiThe Best Mainstream Graphics Card If You Want to Spend a Bit MoreNvidia GeForce RTX 5070 TiThe Nvidia GeForce RTX 5070 Ti is a mid-to-high end graphics card that is more than capable of 4K gaming.See it at NeweggProduct SpecificationsCUDA Cores/Stream Processors8,960Base Clock2.30GHzBoost Clock2.45GHzVideo Memory16GB GDDR7Memory Bandwidth896 GB/sMemory Bus256-bitPower Connectors1 x 12V-6x6 16-pinOutputs1 x HDMI 2.1b, 3 x DisplayPort 2.1bPROSExcellent 4K performance without paying RTX 5080 moneyHas DLSS 4 MFG for high-refresh monitorsCONSContinues Blackwell's lackluster generational improvementsThe Nvidia GeForce RTX 5070 Ti is a lesson in pricing. At its suggested price of $749, it is one of the best bang-for-your-buck graphics cards of this entire generation. However, with the graphics card just having launched recently, it remains to be seen how available it will be at that suggested price. It is distinctly possible that the only RTX 5070 Ti graphics cards that are available will cost much more than the suggested price, which will make it much harder to recommend. Nvidia GeForce RTX 5070 Ti BenchmarksIf you can find the RTX 5070 Ti at $749, or even up to $800, it is one of the best 4K graphics cards for most people. When I reviewed the 5070 Ti, I found that it stayed within reaching distance of the RTX 5080, consistently sitting just 13-15% behind it, while having a 33% lower asking price. It's not hard to see why it's the best value high-end graphics card of this generation, so far at least. The elephant in the room, however, is Blackwell's tepid generation-on-generation improvement over the RTX 4000 graphics cards. The Nvidia RTX 5070 Ti has the best generational gain out of any of the RTX 5000 cards, and its still limited to being just 11% faster than the RTX 4070 Super and 21% faster than the original RTX 4070. That's not a huge jump, but its still a bigger improvement than the RTX 5080, which is just 15% faster than the RTX 4080 in the same test suite. Nvidia GeForce RTX 5090 Photos3. Nvidia GeForce RTX 5090The Best Nvidia Graphics CardGigabyte Nvidia GeForce RTX 5090The Nvidia GeForce RTX 5090 is now the most powerful graphics card on the market, even if it claims that title with a bit less force than in previous generations. See it at NeweggProduct SpecificationsCUDA Cores/Stream Processors21,760Base Clock2.01GHzBoost Clock2.41GHzVideo Memory32GB GDDR7Memory Bandwidth1,792 GB/sMemory Bus512-bitPower Connectors1 x 16-pinOutputs1 x HDMI 2.1b, 3 x DisplayPort 2.1bSize (Founders Edition)11.9 x 5.39 x 1.9 inches (L x W x H) (Dual Slot)PROSThe most powerful consumer GPU out there, period. DLSS 4 multi frame gen will help further boost frame rates.CONSGeneration-on-Generation improvement not very excitingThere's no way around it, the Nvidia GeForce RTX 5090 is the most powerful graphics card on the market right now. Full stop. While it doesn't mark the same kind of generational growth that the RTX 4090 or even the RTX 3090 did, there's no getting around the fact that it provides the best gaming performance you can get right now especially when you take DLSS multi-frame generation into account. Not only is the RTX 5090 bigger than the 4090, now coming with 21,760 CUDA cores and 32GB of GDDR7 memory, but it also has a much higher power budget. When I reviewed the RTX 5090, I found the next-gen graphics flagship would peak at a staggering 578W, a huge increase from the 448W of the RTX 4090. With all that extra power, Nvidia needed to find a better way to dissipate heat, and introduced a new cooler for its Founders Edition. Instead of doubling down on the triple-fan design of the last couple of generations, Team Green actually found a way to slim the design back down to a dual slot cooler, something I haven't seen in a flagship Nvidia graphics card since the RTX 2080 Ti. Nvidia GeForce RTX 5090 BenchmarksThe company was able to do this by shrinking down the circuit board where the GPU is located, placing it at the center of the card. Each side of that PCB is bookended by pass-through heatsinks, where fans pull cool air through the bottom of the card, and shoot it straight through to the top of your PC case. The thermal engineering of it all is a bit more complicated than that, but even with all that extra power, I only ever saw the RTX 5090 reach up to 87C. That's a high temperature, but its still low enough to game at full blast. On average, I found the RTX 5090 to be about 26% faster than the RTX 4090, when looking at 4K games and synthetic 3DMark benchmarks. That number does diminish at lower resolutions, of course, and you really shouldn't be forking over the $1,999 (or more!) for this GPU if you're not going to play at 4K. Even at 4K, there were certain games that simply didn't have much of a performance uplift due to the limitations of the CPU and I used the Ryzen 7 9800X3D. For better or worse, then, the Nvidia GeForce RTX 5090 shines in the exact kind of workload it's designed for. No holds barred, everything maxed out at 4K, and with minimal upscaling. Gone are the days where DLSS on 'Performance' mode is necessary at 4K at least for this beastly GPU. AMD Radeon RX 7900 XT Photos4. AMD Radeon RX 7900 XTX The Best AMD Graphics CardGigabyte AMD Radeon RX 7900 XTXThis Radeon RX 7900 XTX from Gigabyte keeps the price down a bit, while providing an excellent triple-fan cooler.See it at AmazonProduct SpecificationsCUDA Cores/Stream Processors6,144Base Clock1,929MHzBoost Clock2,365MHzVideo Memory24GB GDDR6Memory Bandwidth960 GB/sMemory Bus384-bitPower Connectors2 x 8-pinOutputs1 x HDMI 2.1a, 2 x DisplayPort 2.1, 1 x USB-CSize11.3 x 4.3 x 2 inches (L x W x H) (Dual Slot)PROSExcellent 4K performanceCompetitive pricing with RTX 4080CONSCan lag behind in some games with ray tracingWhile AMD doesnt really have a GPU that can go blow-for-blow with the RTX 4090, it does have one that is extremely competitive with the RTX 4080 Super, Nvidias next-best and more attainable card: The AMD Radeon RX 7900 XTX. Our hands-on testing of the Radeon RX 7900 XTX made it clear that this GPU is an absolute beast at 4K that can keep up with Nvidia, even beating Team Green in some games. Even in games that are traditionally hard on AMD graphics cards, the Radeon RX 7900 XTX can handle at 4K. Just take a look at Cyberpunk 2077 its able to manage an average of 58 fps at 4K with the ray tracing ultra preset, with FSR enabled. Obviously thats not quite as good as the RTX 4080 in the same test, but frankly it was never going to be. However, in games with lighter ray tracing loads, the Radeon RX 7900 XTX fares a lot better. For instance, in Forza Horizon 5 with everything cranked, the Radeon RX 7900 XTX manages an impressive 158fps at 4K, basically matching the RTX 4080 Super at 159fps. Likewise, in Far Cry 6, the Radeon RX 7900 XTX can deliver an incredible 154fps average, beating out the RTX 4080 Super, which can do 151fps. Like with any graphics card, its important to know what kind of games you want to play. But if youre playing a lot of games with little to no ray tracing, the AMD Radeon RX 7900 XTX is an extremely potent 4K graphics card. Plus with DisplayPort 2.1, its better equipped for higher resolution ultrawide monitors.AMD Radeon RX 7800 XT Benchmarks5. AMD Radeon RX 7700 XTThe Best GPU for 1440pGigabyte AMD Radeon RX 7700 XTPacked with a triple-fan cooler, this massive RX 7700 XT from Gigabyte will keep your GPU cool while it pumps out frames.See it at AmazonProduct SpecificationsCUDA Cores/Stream Processors3,456Base Clock1,435MHzBoost Clock2,171MHzVideo Memory12GB GDDR6Memory Bandwidth432 GB/sMemory Bus192-bitPower Connectors2 x 8-pinOutputs1 x HDMI 2.1a, 2 x DisplayPort 2.1, 1 x USB-CSize10.5 x 5.3 x 2 inches (L x W x H) (Dual Slot)PROSExcellent performance for the moneyGreat at 1440pCONSLags behind in ray tracing gamesWhen you think of the perfect resolution for games, its easy to jump to 4K and say more pixels is better." However, 1440p really is the sweet spot for PC gaming, and the Radeon RX 7700 XT is the star of the show, thanks to the performance you get for the money. AMD has lowered the price on the Radeon RX 7700 XT from its launch price of $449 down to $419, but you can easily find it for around $399 on Amazon. At this price, it goes head to head against the Nvidia GeForce RTX 4060 Ti, and thats not a fight that looks good for Nvidia. And while I didn't do a standalone review for the 7700 XT, I tested it alongside the Radeon RX 7800 XT, its big sibling. While the RTX 4060 Ti does beat out the 7700 XT in Cyberpunk 2077, its not by much. The RTX 4060 Ti gets 62fps at 1440p, and the Radeon RX 7700 XT gets 49fps. And thats on the ray tracing ultra preset. Nvidias lead disappears in other games, though. For instance, in Forza Horizon 5, maxed out at 1440p, the RX 7700 XT manages 118fps, compared to 107 from the RTX 4060 Ti. Likewise, in Total War: Warhammer 3, which doesnt have any fancy ray tracing tech, the RTX 4060 Ti gets smoked, with it scoring 68 fps to AMDs 85. Thats a 20% lead at about the same price. There is a catch, though. While the Radeon RX 7700 XT is more powerful than the RTX 4060 Ti, its also more power-hungry, drawing up to 312W from the wall in my testing. Compared to the 159W that the RTX 4060 Ti eats up, thats a pretty huge gap. However, if you have a power supply that can handle it AMD recommends 550W you get a much more powerful card at the same price. Whether or not it affects your power bill down the line is a future issue. Nvidia RTX 40606. Nvidia GeForce RTX 4060The Best GPU for 1080pMSI Nvidia GeForce RTX 4060This RTX 4060 from MSI keeps the price under $300, while delivering a solid cooler. Plus, it should fit even in the smallest gaming rigs.See it at AmazonProduct SpecificationsCUDA Cores/Stream Processors3,072Base Clock1,830MHzBoost Clock2,460MHzVideo Memory8GB GDDR6Memory Bandwidth272 GB/sMemory Bus128-bitPower Connectors1 x 12-pinOutputs1 x HDMI 2.1, 3 x DisplayPort 1.4 Size9.4 x 4.4 x 1.6 inches (L x W x H) (Dual Slot)PROSAffordableSolid 1080p performance with ray tracingCONSOnly 8GB of memoryAccording to the Steam Hardware Survey, 1080p gaming displays are still the most popular monitors out there, and it's not close. So, despite how hyped up 4K displays are, if youre comfortable with 1080p, you dont need an extremely powerful graphics card to play even the most demanding games. Thats where the Nvidia GeForce RTX 4060 really shines. The RTX 4060 can power any game at 1080p, even with all the fancy ray tracing settings that are coming out these days. And thanks to DLSS, it can even stretch to higher resolutions in certain games. And for a current-generation graphics card that you can find under $300, thats not a bad deal. When I reviewed the RTX 4060, I found that it was able to run most games extremely well at 1080p with all the settings cranked at more than 60fps, even if it just barely got there in some games. For instance, in Cyberpunk 2077 on the ray tracing ultra preset, the RTX 4060 scored just 62fps at 1080p. That is the golden frame rate, but youre not going to be doing much high-refresh gaming on this card. RTX 4060 benchmarksThe RTX 4060 does have a weak side, though, and thats how it compares to the RTX 3060 Ti that came out before it. You can still find the RTX 3060 Ti at most retailers, and for a comparable price, but yet it manages to beat out the RTX 4060 in most of my tests. There are some features youd be giving up by going with a last-generation card, DLSS 3.0 being chief among them, but that may be worth it for the slight edge in performance you can get. However, at this price range, a feature like DLSS 3.0 that can stretch performance out in supported games is worth it, and can really stretch out the value of the card. The Nvidia GeForce RTX 4060 isnt perfect, but if you want a current-generation graphics card for under $300 that runs any PC game above 60fps, Id have a hard time coming up with an alternative certainly not the Radeon RX 7600. Upcoming GPUsNow that the latest generation of Nvidia graphics cards is out, starting with the RTX 5090 and RTX 5080, 2025 seems to be a packed year for graphics cards. Team Green followed up its flagship-level cards with the RTX 5070 Ti, which really helps bridge the gap between the mid-range and the high end. However, because this is still an expensive 4K card, anyone looking for 1440p gaming is probably better off waiting for either the Nvidia RTX 5070 or AMD's upcoming cards. Because Nvidia isn't alone in new graphics cards. AMD's Radeon RX 9070 and RX 9070 XT are right around the corner, launching in March 2025. I got a chance to sneak in a benchmark of the 9070 at AMD's CES 2025 booth, and while those are preliminary results you should take with a heaping of salt, it looks like it's going to give the Radeon RX 7900 XT and RTX 4080 a run for their money. Best Graphics Cards FAQAMD or Nvidia? Or Intel?When it comes to which brand of graphics card you should get, it ultimately comes down to your personal preference even if each brand has its own unique advantages. Intel graphics cards are the most affordable option on the market right now, but its graphics cards arent exactly the fastest in the land. On the other hand, Nvidia makes the most powerful GPUs around, but youre going to have to pay out the nose for that performance.AMD graphics cards strike a pretty good balance between the two, but while the company loves to use open-source graphics APIs, it means you dont get access to some of the exclusive features that Nvidia owners enjoy features like DLSS (and the new DLSS 4). AMD has alternatives for literally every Nvidia software feature, but some of them simply arent as good. See our guide to AMD vs. Nvidia GPUs for more info.What power supply should I get?Graphics cards, especially high-end ones, are sucking up more electricity with every passing year. If youre looking to build a new gaming PC, or even upgrade from an older graphics card, you should really consider upgrading to one of the best power supplies.Some of the graphics cards out there right now can take upwards of 450W of power by themselves, so you may want to consider a 1,000W power supply especially if youre going for the Nvidia GeForce RTX 4090.GTX vs. RTXNvidia has both an RTX, or Ray Tracing Texel eXtreme, and GTX, Giga Texel Shader eXtreme graphics cards series, with the RTX offerings being newer, more powerful, and more expensive. That boosted performance is thanks to the architecture of RTX cards, which offer both Tensor and RT cores alongside CUDA cores for better graphics and rendering. Tensor cores enable AI and high-performance computing tasks bringing support for DLSS tech to help with upscaling and sharpening. RT cores are dedicated to ray tracing, allowing for more realistic lighting and shadows in scenes. Nvidias GTX graphics cards have a much simpler architecture and dont offer Tensor or RT cores. Though they still work for budget builds, GTX cards will soon be obsolete. Where to Get the Best Graphics Cards in the UKThere aren't too many differences when it comes to the graphics cards you can pick up in the UK, but the main takeaway is where you can purchase them. All of the following links have been updated with UK vendors, saving you some time and money if you're interested in picking up any of the graphics cards we've mentioned. Best Graphics Card Asus TUF Gaming RTX 4070 Ti OC EditionSee it at Currys PC WorldBest Budget Graphics Card MSI Geforce RTX 3050 Gaming XSee it at AmazonBest AMD Graphics Card XFX Speedster MERC310 RX 7900XTSee it at AmazonBest 4K Graphics Card Nvidia GeForce RTX 4080 See it at NvidiaJacqueline Thomas is the Hardware and Buying Guides Editor at IGN. When she's not helping her friends and family buy computers, you can usually find her tinkering with her own PC.

In his own words, screenwriter Steven Knight does legends. The Peaky Blinders, Taboo, and SAS: Rogue Heroes creator takes elements of real-life history and turns them into swaggering myths.With an ear for an unusual name, and a knack for conjuring flesh-and-blood icons out of historical figures that crop up in newspaper reports and census columns, Knight has rehabilitated the British working class period drama. Hes taken the gratitude and drudgery out, and written ambition, glamour and modernity in. He made icons out of real-life Birmingham gangsters in Peaky Blinders, and hes about to do the same with East-End boxers and female thieves The Forty Elephants in new drama A Thousand Blows.As the six-episode first series (a second has already been filmed) arrives on Disney+ in the UK and Hulu in the US, lets have an overview of the real history of A Thousand Blows, and some pointers on where you can find out more.Hezekiah Moscow, Alec Munroe & Sugar Goodson Were All RealTo learn about the real Hezekiah, Alec, Sugar, Treacle and more, go no further than the historical research that inspired A Thousand Blows, conducted by the shows boxing historian and historical consultant Sarah Elizabeth Cox. First published online in 2019 on her Grappling With History website and currently being expanded into a book, Coxs findings include photographs, posters, newspaper articles, census entries and more detailing the lives and careers of the real people who inspired the shows characters. As Cox writes, A Thousand Blows is not a documentary: the characters and storylines are the creation of a wonderful team of writers, and it is only in little snippets here and there that they cross paths with reality.Those crossed-paths with reality include the real existence of a West-Indian immigrant named Hezekiah Moscow, who worked at the East London Aquarium as a bear and lion tamer and competed in various boxing bouts, including at the real Blue Coat Boy pub in Shoreditch, London. His West-Indian trainer and corner man Alec Munroe was also real. There was indeed an East-End fighter nicknamed Sugar Goodson though according to Cox, the real Sugar was thought to have only one eye.Mary Carr and The Forty Elephants Were a Real Criminal GangSpeaking to the BBC, A Thousand Blows creator Steven Knight explained how he combined two real-life stories to create the Disney+/Hulu drama. Approached by actor-producers Hannah Walters and Stephen Graham about putting the life of boxer Hezekiah Moscow on screen with their company Matriarch Productions, Knight combined that story with another based in historical fact hed been wanting to tell about female thief gang The Forty Elephants.A story about a real person who came from Jamaica with an ambition to become a lion tamer and became a really famous boxer? Thats pretty much irresistible.And when I dug into it and found out about this person and his experiences, it was very compelling. Before then, for a long time, Id wanted to tell the story of the Forty Elephants. Both of those true stories are amazing, and the fact is they were both happening at the same time and in the same place. I thought it would be interesting to imagine what would have happened if Mary and Hezekiah had met and thats what this show is about.Mary Carr was indeed the Queen of the Forty Elephants around the 1880s period in which A Thousand Blows is set, as well as being an artists model for painter Frederic Leighton. Read more about the Elephant and Castle-based gangs tactics and lifestyle via the BBC here.Mild Spoiler warning: references to plot details in A Thousand Blows below.The Morant Bay Rebellion Was an Infamous Part of Jamaican HistoryHezekiahs traumatic flashback memories to his childhood in Jamaica elliptically tell a version of a real, violent historical episode in colonial history. The 1865 Morant Bay Rebellion was an uprising by the people of Jamaicas southeast coast against cruel privations by their British colonial oppressors. You can read more about its origins and impact at the national archives website here.Join our mailing listGet the best of Den of Geek delivered right to your inbox!Li Hongzhang and Lo Feng Luh Were Real Chinese DiplomatsThe Chinese dignitaries who are guests of the Earl of Lonsdale in A Thousand Blows are based on real diplomats who served as part of a legation to London in the aftermath of the 19th century Opium Wars between Britain and China. You can see here an artists drawing of Lo Feng Luh (played by Chike Chan in the show), and here a contemporary newspaper report about the Chinese minister.The 5th Earl of Lonsdale Was a Real Boxing EnthusiastHugh Cecil Lowther was a real English peer and sportsman during the period around which A Thousand Blows is set. Hes the Lonsdale behind the famous British sports brand of the same name, and was a founding member of the National Sporting Club who was said to have donated the very first Lonsdale Boxing Belts for the boxing championship trophy.Aerialist Miss La La Was RealThe acrobat that Mary and Hezekiah see performing at a West End music hall is based on a real Black Polish historical figure, also named Miss La La. She performed on a swing high above crowds and her feats were recorded by painter Edgar Degas in his 1879 work Miss La La at the Cirque Fernando. She was the subject of an exhibition at Londons National Gallery in 2024, which you can read more about here.Queen Victoria Did Have a Black GoddaughterThe A Thousand Blows character Victoria Davies must be inspired by the real Nigerian-born woman known as Sara Sally Forbes Bonetta (originally called Aina, before she was renamed by the English captain to whom she was discharged by enslaved people trader King Ghezo of Dahomey), who became Queen Victorias goddaughter. The timeline doesnt quite match up as the real Aina died young of tuberculosis and didnt survive into 1880 period in which A Thousand Blows is set, but its very close. See portraits of her here, courtesy of the National Portrait Gallery.The Blue Coat Boy Pub, its Boxing Ring, and its Landlord Were RealThe Blue Coat Boy Pub had an MC who owned a boxing saloon nicknamed William Punch Lewis, just like Daniel Mays character in the TV show. Read more about them all here, thanks to the research of Sarah Elizabeth Cox.A Thousand Blows is streaming now on Disney+ in the UK and Hulu in the US.