0 Comentários
0 Compartilhamentos
152 Visualizações
Diretório
Diretório
-
Faça Login para curtir, compartilhar e comentar!
-
WWW.CREATIVEBLOQ.COM'The posters are weird': Inside the unapologetically bold art of Soho Rep theatreIllustration meets the stage in this immersive visual project.0 Comentários 0 Compartilhamentos 147 Visualizações
-
WWW.WIRED.COMNinja Crispi Review: An Ingenious Portable Air FryerThis new Ninja Crispi portable air fryer may be the new best solution to potlucks, office lunch, and maybe tailgating.0 Comentários 0 Compartilhamentos 148 Visualizações
-
WWW.WIRED.COMBookshop.org Now Sells EbooksThe bookseller is applying its sales modelwhere online purchases support indie bookstoresto digital books. It has also released a mobile app for shopping and reading ebooks.0 Comentários 0 Compartilhamentos 141 Visualizações
-
WWW.NYTIMES.COMG.M. Has Plans Ready for Trumps Canada and Mexico TariffsGeneral Motors, the largest producer of cars in Mexico, wont provide details on how it would react if President Trump imposes 25 percent tariffs from the two countries.0 Comentários 0 Compartilhamentos 127 Visualizações
-
WWW.MACWORLD.COMThe latest iPhone update patches a security flaw exploited since 2023MacworldApple on Monday released a flurry of updates for all of its devices that provide several new features and fix a handful of bugs. But as always, the most important reason to upgrade is because they make your devices safer.Among the numerous security updates, the iOS/iPadOS 18.3, macOS Sequoia 15.3, watchOS 11.3, tvOS 18.3, and visionOS 2.3 updates patch a CoreMedia zero-day flaw. Exploitation of CVE-2025-24085 could allow a malicious application to access privileged parts of the system and was fixed with improved memory management.CoreMedia is a framework used for time-based audio-visual assets such as podcast apps and other media players.Apple says it is aware of a report thatthe vulnerability may have been actively exploited using versions of iOS before iOS 17.2, which arrived in December 2023. Interestingly, Apple didnt release iOS 17.7.3 with a fix for phones not running iOS 18, though all phones that were compatible with iOS 17 are also compatible with iOS 18.For older tablets, Apple released iPadOS 17.7.3, though it does not include the CoreMedia fix. The updates for older Macs (macOS Sonoma 14.7.3 and macOS Ventura 13.7.3) also dont include a CoreMedia fix. Among the other security patches in the updates are several AirPlay fixes and a FaceTime fix related to apps accessing user-sensitive data.Apple didnt divulge who discovered the security flaw. Its also unclear how the vulnerability was exploited. You can update your Apple device by going to the Settings app (System Settings on a Mac) and selecting the Software Update tab.0 Comentários 0 Compartilhamentos 130 Visualizações
-
WWW.MACWORLD.COMDeepSeek is making everyone look silly, except AppleMacworldIt is time once again to see how things are going on the good ship Artificial Intelligence, sailing merrily through the seas of current technology.Sorry?Look, the Macalope just writes about tech, he doesnt get to decide whats currently hot. Clearly. Or wed all be talking about the cool new small phones we have.At any rate, no need to worry. It is smooth sailing for our glorious leaders in AI. Everyone loves the product, venture capital funding keeps rolling in and the moats are dee-iPhone users turn on to DeepSeek AIThat sound you heard early Monday morning was not the earthquake in Boston but rather the sound of AI stocks crashing to the ground after the Chinese app DeepSeek was unveiled.By the way, stay in your lane, Boston. Dont try to steal our beloved West Coast institutions like earthquake anxiety. Very rude. Its not like you see us running around inventing new chowders.The reason for the anxiety over DeepSeek is that apparently, the Chinese developers have found a way to engineer an AI that uses a fraction of the processing power and money while still delivering the same laughably incorrect answers as competing models from Google, Microsoft, and ChatGPT.In addition to shares of AI companies sliding, Nvidias shares also tumbled on the fear that it would not be able to personally thaw the ice caps.The Chinese AI startup behind the model was founded by hedge fund manager Liang Wenfeng, who claims they used just 2,048 Nvidia H800s and $5.6 million to train R1 with 671 billion parameters, a fraction of what OpenAI and Google spent to train comparably sized models.Oops. The Macalope supposes they do not get the rarified water that we have here in the good ol you ess of ay that causes the brains of venture capitalists to soften to the point where they shoot money out of a t-shirt canon at anything their buddy Pete told them to aim at.The lack of a moat around these companies was already predicted by lots of people, as early as 2023. Now its starting to look like maybe there wasnt even a wall.(Tell the Macalope again how Apple is the one behind it and was foolish not to invest huge bucks in this.)IDGPart of the problem is that these AIs are learning from each other. When you spend billions and melt Antarctica to come up with a chatty way to deliver a bunch of admittedly useful code suggestions (which you stole from some poor schlub on the internet) and great recipes for rock pizza because your AI cant interpret sarcasm, someone else is going to spend just millions to lift that from you. Again, that would be really sad except you scraped content that you didnt license and laughed off the copyright concerns of the people you stole it from. Sucks to be you.This all, of course, comes after (during?) the whole TikTok debacle which raises the question of whether or not well have to deal with more rigamarole. Presumably, the current president will suggest a ban of or tariffs on or forced deportation of DeepSeek and then the subsequent Hunter Biden administration will enact that ban only to have the Baron Trump administration grandiosely (and possibly illegally) rescind the ban.Meanwhile, the rest of us will be ankle-deep in water and will have forgotten what eggs taste like.If youre not in the U.S., then its merely a choice between whether or not an American company will have your queries or a Chinese company will. Pick your poison.The Macalope knows he just wrote a column two weeks ago in which he tried to show his views of AI are more nuanced than youd think, but while the technology has some very good applications, the companies and business models that surround it can go suck those rocks they tell us to put on our pizzas.The one and only piece of evidence you need for this is OpenAI CEO Sam Altmans recent redefinition of artificial general intelligence. According to the leading company in AI (at least as of the close of business last Friday), its not about the specific capabilities of the system. No, its about being able to put enough regular people out of work in order to generate $100 billion in profit.AI could be a great technology. Its too bad its currently in the hands of a bunch of sentient popped Izod collars. If DeepSeek is here to take some of the air out of their proverbial tires, the Macalope is popping corn, not collars.0 Comentários 0 Compartilhamentos 127 Visualizações
-
WWW.COMPUTERWORLD.COMDeepSeek triggers shock waves for AI giants, but the disruption wont lastChinese start-up DeepSeek rocked the tech industry Monday after the companys new generative AI (genAI) bot hit Apples App Store and Google Play Store and downloads almost immediately exceeded those of OpenAIs ChatGPT. US AI model and chipmaker stock prices were hit hard by the newcomers arrival; Google, Meta and OpenAI all initially suffered and chipmaker Nvidias stock closed the day down 17%. (The tech heavy Nasdaq exchange lost more than 600 points.)DeepSeeks open-source AI models impact lies in matching US models performance at a fraction of the cost by using compute and memory resources more efficiently. But industry analysts believe investor reaction to DeepSeeks impact on US tech firms and others is being dramatically exaggerated.The market is incorrectly presuming this as a zero-sum game, said Chirag Dekate, a vice president analyst at Gartner Research. Theyre basically saying, Maybe we dont need to build data centers anymore, maybe were not as energy starved because DeepSeek showed us we can do more with less.Giuseppe Sette, president of AI tech firm Reflexivity agreed, stressing that DeepSeek took the market by storm by doing more with less.In layman terms, they activate only the most relevant portions of their model for each query, and that saves money and computation power, Sette said. This shows that with AI, the surprises will keep on coming in the next few years. And even though that might be a bit of a shocker today, its extremely bullish in the long-term because it opens the way for deeper and broader adoption of AI at all scales.In essence, the markets have overlooked that companies such as Google, Meta, and OpenAI can replicate DeepSeeks efficiencies with more mature, scalable AI models that offer better security and privacy.This is not a the sky is falling moment for markets. I think they should take a close look at what this actually is: there are techniques you can implement to more effectively scale your AI models, Dekate said.Another looming problem for the newcomer is that DeepSeek is purported to filter out content that could be viewed as critical of the Chinese Communist government. DeepSeeks release of its R1 and R1-Zero reasoning models on Jan. 20 quickly drew attention for two key aspects:DeepSeek eliminates human feedback in training, speeding up model development, according to AI developer Ben Thompson.DeepSeek requires less memory and compute power, needing fewer GPUs to perform the same tasks as other models.DeepSeek claims its breakthroughs in AI efficiency cost less than $6 million and took less than two months to develop.John Belton, a portfolio manager atGabelli Funds, an asset management firm whose funds include shares of Nvidia, Microsoft, Amazon, and others, said DeepSeeks achievements are real, but some of the companys claims are misleading.No, you cannot recreate DeepSeek with $6 million and the extent to which they distilled existing models (took shortcuts potentially without license) is an unknown, Belton said via email to Computerworld. However, they have made key breakthroughs that show how to reduce training and inference costs.Belton also pointed out that DeepSeek isnt new. Its creator, Liang Wenfeng, a hedge fund manager and AI enthusiast, published a paper on the performance breakthroughs more than a month ago and released a model with similar methods a year ago.Dekate said DeepSeeks rollout was particularly timely because just last month news outlets were publishing stories about AI scaling limitations from leading providers.As organizations continue toembrace genAI tools and platforms and explore how they can create efficiencies and boost worker productivity, theyre also grappling with the high costs and complexity of the technology.DeepSeek improved memory bandwidth efficiency with two key innovations: using a lower-position memory algorithm and switching from FP32 (32-bit) to FP8 (8-bit) for model precision training. Theyre using the same amount of memory to store and move more data, Dekate said.One analogy would be to consider the onramp to a major city highway the highway being the data path. If the onramp only has one lane, there are only two ways to address traffic congestion:Increase the width of the roadway to fit more trafficReduce the size of the vehicles to more fit on the roadwayDeepSeek did both. It created smaller vehicles, i.e., it used smaller data packets (8-bit) and therefore was able to pack more data into the same footprint.The second key innovation was optimizing and compressing the key-value cache. DeepSeek used compression algorithms to reduce memory by processing prompts in two phases: decomposing and generating responses, both relying on efficient key-value cache use.They utilized underlying compute and memory resources incredibly efficiently, Dekate said. That is an amazing accomplishment, because theyre utilizing the underlying GPU resources more productively. Their models are able to perform at leadership-class levels while using a relatively lower scale of resources.Enterprises can benefit as well by adopting the techniques introduced by DeepSeek because it reduces the cost of adoption by using fewer compute resources for inferencing and training. Lower model costs should benefit innovators such as OpenAI and reduce the cost of applying AI across industries. By using resources more efficiently, DeepSeek enables faster, broader AI adoption by other companies, driving growth in AI development, demand, and infrastructure.And in the end, DeepSeeks algorithm still needs AI accelerator technology to work meaning GPUs and ASICs.Its not the case that DeepSeek just woke up one day and had an amazing breakthrough. No, theyre using sound engineering techniques and theyre using some of the leading AI accelerators and GPUs happen to be table stakes, Dekate said. And they use thousands of them. Its not like they discovered a new technique that blew this whole space wide open. No. You still need AI accelerators to perform model training.Even in the most pessimistic view, if AI costs drop to 5% of those from other leading AI models, that efficiency eventually benefits those other models by reducing their costs, allowing for faster model adoption.For enterprises, Dekate said, its worth exploring DeepSeek and similar models internally and in private settings. Your legal team evaluates the terms and conditions of your ecosystem quite extensively. Theyll ask if privacy is protected. Are the data sources filtered? Are AI model responses filtered in any sense? he said.Before jumping in, enterprises should carefully consider these details. Models like Gemini and GPT offer reliable, secure responses with enterprise-level protections, unlike many open models that lack these controls, Dekate argued.Once everything settles, the net-net is that DeepSeek has developed very specific capabilities that are quantitative and thats something to learn from, just as they did from Llama 3, Dekate said.0 Comentários 0 Compartilhamentos 142 Visualizações
-
WWW.TECHNOLOGYREVIEW.COMAIs energy obsession just got a reality checkThis story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first,sign up here.Just a week in, the AI sector has already seen its first battle of wits under the new Trump administration. The clash stems from two key pieces of news: the announcement of the Stargate project, which would spend $500 billionmore than the Apollo space programon new AI data centers, and the release of a powerful new model from China. Together, they raise important questions the industry needs to answer about the extent to which the race for more data centerswith their heavy environmental tollis really necessary.A reminder about the first piece: OpenAI, Oracle, SoftBank, and an Abu Dhabibased investment fund called MGX plan to spend up to $500 billion opening massive data centers around the US to build better AI. Much of the groundwork for this project was laid in 2024, when OpenAI increased its lobbying spending sevenfold (which we were first to report last week) and AI companies started pushing for policies that were less about controlling problems like deepfakes and misinformation, and more about securing more energy. Still, Trump received credit for it from tech leaders when he announced the effort on his second day in office. I think this will be the most important project of this era, OpenAIs Sam Altman said at the launch event, adding, We wouldnt be able to do this without you, Mr. President.Its an incredible sum, just slightly less than the inflation-adjusted cost of building the US highway system over the course of more than 30 years. However, not everyone sees Stargate as having the same public benefit. Environmental groups say it could strain local grids and further drive up the cost of energy for the rest of us, who arent guzzling it to train and deploy AI models. Previous research has also shown that data centers tend to be built in areas that use much more carbon-intensive sources of energy, like coal, than the national average. Its not clear how much, if at all, Stargate will rely on renewable energy.Even louder critics of Stargate, though, include Elon Musk. None of Musks companies are involved in the project, and he has attempted to publicly sow doubt that OpenAI and SoftBank have enough of the money needed for the plan anyway, claims that Altman disputed on X. Musks decision to publicly criticize the presidents initiative has irked people in Trumps orbit, Politico reports, but its not clear if those people have expressed that to Musk directly.On to the second piece. On the day Trump was inaugurated, a Chinese startup released an AI model that started making a whole bunch of important people in Silicon Valley very worried about their competition. (This close timing is almost certainly not an accident.)The model, called DeepSeek R1, is a reasoning model. These types of models are designed to excel at math, logic, pattern-finding, and decision-making. DeepSeek proved it could reason through complicated problems as well as one of OpenAIs reasoning models, o1and more efficiently. Whats more, DeepSeek isnt a super-secret project kept behind lock and key like OpenAIs. It was released for all to see.DeepSeek was released as the US has made outcompeting China in the AI race a top priority. This goal was a driving force behind the 2022 CHIPS Act to make more chips domestically. Its influenced the position of tech companies like OpenAI, which has embraced lending its models to national security work and has partnered with the defense-tech company Anduril to help the military take down drones. Its led to export controls that limit what types of chips Nvidia can sell to China. The success of DeepSeek signals that these efforts arent working as well as AI leaders in the US would like (though its worth noting that the impact of export controls for chips isnt felt for a few years, so the policy wouldnt be expected to have prevented a model like DeepSeek).Still, the model poses a threat to the bottom line of certain players in Big Tech. Why pay for an expensive model from OpenAI when you can get access to DeepSeek for free? Even other makers of open-source models, especially Meta, are panicking about the competition, according to The Information. The company has set up a number of war rooms to figure out how DeepSeek was made so efficient. (A couple of days after the Stargate announcement, Meta said it would increase its own capital investments by 70% to build more AI infrastructure.)What does this all mean for the Stargate project? Lets think about why OpenAI and its partners are willing to spend $500 billion on data centers to begin with. They believe that AI in its various formsnot just chatbots or generative video or even new AI agents, but also developments yet to be unveiledwill be the most lucrative tool humanity has ever built. They also believe that access to powerful chips inside massive data centers is the key to getting there.DeepSeek poked some holes in that approach. It didnt train on yet-unreleased chips that are light-years ahead. It didnt, to our knowledge, require the eye-watering amounts of computing power and energy behind the models from US companies that have made headlines. Its designers made clever decisions in the name of efficiency.In theory, it could make a project like Stargate seem less urgent and less necessary. If, in dissecting DeepSeek, AI companies discover some lessons about how to make models use existing resources more effectively, perhaps constructing more and more data centers wont be the only winning formula for better AI. That would be welcome to the many people affected by the problems data centers can bring, like lots of emissions, the loss of fresh, drinkable water used to cool them, and the strain on local power grids.Thus far, DeepSeek doesnt seem to have sparked such a change in approach. OpenAI researcher Noam Brown wrote on X, I have no doubt that with even more compute it would be an even more powerful model.If his logic wins out, the players with the most computing power will win, and getting it is apparently worth at least $500 billion to AIs biggest companies. But lets rememberannouncing it is the easiest part.Now read the rest of The AlgorithmDeeper LearningWhats next for robotsMany of the big questions about AI-how it learns, how well it works, and where it should be deployedare now applicable to robotics. In the year ahead, we will see humanoid robots being put to the test in warehouses and factories, robots learning in simulated worlds, and a rapid increase in the militarys adoption of autonomous drones, submarines, and more.Why it matters: Jensen Huang, the highly influential CEO of the chipmaker Nvidia, stated last month that the next advancement in AI will mean giving the technology a body of sorts in the physical world. This will come in the form of advanced robotics. Even with the caveat that robotics is full of futuristic promises that usually arent fulfilled by their deadlines, the marrying of AI methods with new advancements in robots means the field is changing quickly. Read more here.Bits and BytesLeaked documents expose deep ties between Israeli army and MicrosoftSince the attacks of October 7, the Israeli military has relied heavily on cloud and AI services from Microsoft and its partner OpenAI, and the tech giants staff has embedded with different units to support rollout, a joint investigation reveals. (+972 Magazine)The tech arsenal that could power Trumps immigration crackdownThe effort by federal agencies to acquire powerful technology to identify and track migrants has been unfolding for years across multiple administrations. These technologies may be called upon more directly under President Trump. (The New York Times)OpenAI launches Operatoran agent that can use a computer for youOperator is a web app that can carry out simple online tasks in a browser, such as booking concert tickets or making an online grocery order. (MIT Technology Review)The second wave of AI coding is hereA string of startups are racing to build models that can produce better and better software. But its not only AIs increasingly powerful ability to write code thats impressive. They claim its the shortest path to superintelligent AI. (MIT Technology Review)0 Comentários 0 Compartilhamentos 145 Visualizações
-
APPLEINSIDER.COMJudge tells Apple it's too late to intervene over Google monopoly caseGoogle has been found in court to be a monopoly, but there is to be a trial to plan how to remedy the situation, and the judge has now refused to accept Apple intervening, saying it waited too long.Google is the default search engine on iPhone because it pays to beThe US Department of Justice's (DOJ) antitrust lawsuit against Alphabet, the parent company of Google, reached a key point in August 2024. Judge Amit Metha ruled that Google is a search and advertising monopoly, and is in breach of Section 2 of the Sherman Act.Following this, the trial continued, and it is shortly to be followed by what's called a remedy trial. According to Courthouse News, Apple filed a motion to defend its "property interest" over its contract worth billions of dollars with Google over search on the iPhone and it's been dismissed. Continue Reading on AppleInsider | Discuss on our Forums0 Comentários 0 Compartilhamentos 135 Visualizações