The company's parent firm Recruit's CEO, Hisayuki Idekoba, will return to his post as CEO of Indeed.

This isn't a blip on the radar - learn how AI is changing search.

In context: TSMC has steadily raised the prices of its most advanced semiconductor process nodes over the past several years – so much so that one analysis suggests the cost per transistor hasn't decreased in over a decade. Further price hikes, driven by tariffs and rising development costs, are reinforcing the notion that Moore's Law is truly dead. The Commercial Times reports that TSMC's upcoming N2 2nm semiconductors will cost $30,000 per wafer, a roughly 66% increase over the company's 3nm chips. Future nodes are expected to be even more expensive and likely reserved for the largest manufacturers. TSMC has justified these price increases by citing the massive cost of building 2nm fabrication plants, which can reach up to $725 million. According to United Daily News, major players such as Apple, AMD, Qualcomm, Broadcom, and Nvidia are expected to place orders before the end of the year despite the higher prices, potentially bringing TSMC's 2nm Arizona fab to full capacity. Also see: How profitable are TSMC's nodes: crunching the numbers Unsurprisingly, Apple is getting first dibs. The A20 processor in next year's iPhone 18 Pro is expected to be the first chip based on TSMC's N2 process. Intel's Nova Lake processors, targeting desktops and possibly high-end laptops, are also slated to use N2 and are expected to launch next year. Earlier reports indicated that yield rates for TSMC's 2nm process reached 60% last year and have since improved. New data suggests that 256Mb SRAM yield rates now exceed 90%. Trial production is likely already underway, with mass production scheduled to begin later this year. // Related Stories With tape-outs for 2nm-based designs surpassing previous nodes at the same development stage, TSMC aims to produce tens of thousands of wafers by the end of 2025. TSMC also plans to follow N2 with N2P and N2X in the second half of next year. N2P is expected to offer an 18% performance boost over N3E at the same power level and 36% greater energy efficiency at the same speed, along with significantly higher logic density. N2X, slated for mass production in 2027, will increase maximum clock frequencies by 10%. As semiconductor geometries continue to shrink, power leakage becomes a major concern. TSMC's 2nm nodes will address this issue with gate-all-around (GAA) transistor architectures, enabling more precise control of electrical currents. Beyond 2nm lies the Angstrom era, where TSMC will implement backside power delivery to further enhance performance. Future process nodes like A16 (1.6nm) and A14 (1.4nm) could cost up to $45,000 per wafer. Meanwhile, Intel is aiming to outpace TSMC's roadmap. The company recently began risk production of its A18 node, which also features gate-all-around and backside power delivery. These chips are expected to debut later this year in Intel's upcoming laptop CPUs, codenamed Panther Lake.

This audio is auto-generated. Please let us know if you have feedback. Julianne Malveaux is an MIT-educated economist, author, educator and political commentator who has written extensively about the critical relationship between public policy, corporate accountability and social equity.  The rapid expansion of data centers across the U.S. is not only reshaping the digital economy but also threatening to overwhelm our energy infrastructure. These data centers aren’t just heavy on processing power — they’re heavy on our shared energy infrastructure. For Americans, this could mean serious sticker shock when it comes to their energy bills. Across the country, many households are already feeling the pinch as utilities ramp up investments in costly new infrastructure to power these data centers. With costs almost certain to rise as more data centers come online, state policymakers and energy companies must act now to protect consumers. We need new policies that ensure the cost of these projects is carried by the wealthy big tech companies that profit from them, not by regular energy consumers such as family households and small businesses. According to an analysis from consulting firm Bain & Co., data centers could require more than $2 trillion in new energy resources globally, with U.S. demand alone potentially outpacing supply in the next few years. This unprecedented growth is fueled by the expansion of generative AI, cloud computing and other tech innovations that require massive computing power. Bain’s analysis warns that, to meet this energy demand, U.S. utilities may need to boost annual generation capacity by as much as 26% by 2028 — a staggering jump compared to the 5% yearly increases of the past two decades. This poses a threat to energy affordability and reliability for millions of Americans. Bain’s research estimates that capital investments required to meet data center needs could incrementally raise consumer bills by 1% each year through 2032. That increase may seem small at first, but it can add up quickly for households already struggling with high energy prices. As utilities attempt to pay for these upgrades, the burden could fall on consumers' shoulders unless policies are enacted to make the tech companies driving this demand handle the costs. One example comes from Ohio, where the boom in data centers means central Ohio is on track to use as much power as Manhattan by 2030. There, the state's largest energy company, American Electric Power, has proposed a new rate structure for data centers that requires them to pay at least 85% of their predicted energy demand every month, even if they use less, to ensure the utility won't need to pass off costs for expanded infrastructure to consumers. States could also consider passing legislation to impose a temporary tax on new high-usage energy consumers, like data centers, cryptocurrency miners and chip manufacturing facilities. Those tax dollars could go directly into an energy relief fund, which could be used to offset increased energy costs for current consumers, either through a tax rebate or by funding the construction and maintenance of new infrastructure, so those costs aren't passed down to consumers in the first place. There's opportunity here, too, for policymakers, utilities and data centers to join forces and help drive the clean energy revolution. Policymakers could provide incentives for data centers that adopt energy-saving measures or include renewable energy sources to offset the burden on utilities and consumers. By encouraging tech companies to produce a certain percentage of their own energy on-site, states can reduce the need for costly grid expansions while promoting green energy initiatives. Tech companies have already pushed back against efforts to implement such policies, with a coalition of data center backers that includes Amazon, Microsoft and Meta claiming in Ohio that requiring them to pay higher rates is discriminatory and unprecedented, and that it could discourage future investment in Ohio. The reality, however, is that these tech companies can and should carry the burden of the new energy infrastructure they’re demanding. Amazon’s net earnings for 2023 were $30.4 billion. Microsoft brought in $72.4 billion. Meta? $39 billion. Passing on a fraction of these profits to fund the infrastructure that drives this wealth is a small price to pay to ensure fair treatment of energy consumers. The massive energy demand created by these new data centers is unprecedented. And that’s exactly why it’s important for policymakers and utilities to take action now, and set a precedent that protects average consumers by requiring tech companies to pay their fair share for the electricity they need. If left unaddressed, the unchecked growth of data centers will continue to threaten energy security and affordability for millions of Americans. States and energy companies must adopt policies to prevent the burden of rising electricity demands and prices from falling disproportionately on everyday energy consumers. By ensuring that tech companies contribute fairly to the infrastructure that sustains them, we can build a more sustainable and equitable energy future for all.

There’s a 104 days of summer vacation, but for some folks, school never really ends. At CCXP Mexico on Saturday, Prime Video unveiled its first look for season two of Gen V, the college-focused spinoff for The Boys, and the kids at Godolkin University have their work cut out for them. Set after The Boys’ fourth season amid Homelander effectively ruling the United States, complete with Godolkin students leaving campus through specific checkpoints and forced to wear ID badges, Marie (Jaz Sinclair) and her friends are released from a rehabilitation center and allowed back at school. There, the new dean Cipher (Hamish Linklater) is looking to get students battle ready for a potential Supes/Human war. While the regular Supe kids are content to prepare and party, there’s division among Marie and her friends: some want to run and never look back, others want to stay and tear things apart from the inside. Too bad for them, those plans are interrupted by the discovery of a secret program from the school’s early days which may somehow link to Marie. Shortly before production on the new season began in spring 2024, Chance Perdromo (who played Andre, one of the kids imprisoned with Marie) died in a motorcycle crash. The Gen V team later stated they wouldn’t recast Andre and instead tweaked the season around his passing, and his father Polarity (Sean Patrick Thomas) will be mourning (and later investigating) Andre’s passing. Also starring Derek Luh, Lizzie Broadway, London Thor, and Asa Germann, season two of Gen V kicks off September 17 with a three-episode premiere, followed by a single episode every week through October 22. Want more io9 news? Check out when to expect the latest Marvel, Star Wars, and Star Trek releases, what’s next for the DC Universe on film and TV, and everything you need to know about the future of Doctor Who.

Fans slam Gearbox CEO Randy Pitchford's message to "cost sensitive" fans "We are out here struggling, Randy." Image credit: Eurogamer/Gearbox News by Vikki Blake Contributor Published on June 1, 2025 Gearbox CEO Randy Pitchford has double-doubled down on comments about the cost of Borderlands 4, by inviting "cost sensitive" fans to download Tiny Tina's Wonderlands from the Epic Games Store for "FREE". These latest comments come after Gearbox CEO Randy Pitchford made headlines when he told a X/Twitter commenter that "real fans" would find a way to get their hands on a copy of the game, even if it came with a $80 price tag. He later tried to justify it, all the while saying pricing was "not [his] call". In this latest social media post, Pitchford wrote the word "free" four times, each time in blockcaps. 9 Exciting 2025 Open-World Games We Can't Wait to Play.Watch on YouTube "For our real fans who may be cost sensitive, the very awesome and incredibly fun smash hit videogame Tiny Tina’s Wonderlands is FREE this week on the Epic Games Store," Pitchford wrote. "Please enjoy this FREE gift by grabbing your FREE copy here, FREE." Aforementioned real fans were quick to respond to this comment, too. "You don't get to just walk back being incredibly tone-deaf to the world around you. We are out here struggling, Randy," said one player, while another wrote: "Dude you need to shut up already. No reason games should be $80 especially if its anything like 3 or that shit ass excuse of a movie." "I'm afraid the industry is moving in that direction and it’s just reality that we'll have to accept," Pitchford insisted. "The price for Borderlands 4 is going to get announced by the publisher soon. My wish, having worked my ass off on the game, is for as many people as possible to get to play it." When another commented: "So Randy, you think making backhanded comments towards your consumers about them being 'cost sensitive' is the plan? Wtf happened to you," he responded: "Backhanded? There are literally people who want to play great video games, have a game-ready PC, but may not be able to spend enough to buy a new AAA game for awhile and this week there’s a free offer for one of the best shooter looters to come out over the last five years…" "Saying 'cost sensitive' as the CEO of a company (In lieu of criticism of an $80 price tag for BL4) when a lot of your fan base is struggling to even pay for the roof over their heads and food on their tables is a crazy sentence to utter," said another player. Just a week before the "real fans" kerfuffle, Pitchford implored we "play the game first and understand the choices [the development team has] made" before passing judgement on changes Gearbox has made to the shooter's head's up display, including the removal of the mini-map. Borderlands 4, as we learned a couple of weeks ago, is now arriving earlier than expected on 12th September, and developer Gearbox recently shared more of its latest looter shooter in a new PlayStation showcase. It showed off a couple of new Vault Hunters, some of its new weapons, new traversal mechanics, its new planet, and more all to whip-up enthusiasm for what creative director Graeme Timmins called "hands-down our best Borderlands ever".

Large Reasoning Models (LRMs), trained from LLMs using reinforcement learning (RL), demonstrated great performance in complex reasoning tasks, including mathematics, STEM, and coding. However, existing LRMs face challenges in completing various puzzle tasks that require purely logical reasoning skills, which are easy and obvious for humans. Current methods targeting puzzles focus only on designing benchmarks for evaluation, lacking the training methods and resources for modern LLMs to tackle this challenge. Current puzzle datasets lack diversity and scalability, covering limited puzzle types with little control over generation or difficulty. Moreover, due to the success of the “LLM+RLVR” paradigm, it has become crucial to obtain large, diverse, and challenging sets of verifiable puzzle prompts for training agents. Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a key method for improving models’ reasoning capabilities, removing the need for reward models by directly assigning rewards based on objectively verifiable answers. Puzzles are particularly well-suited for RLVR. However, most prior RLVR research has overlooked the puzzles’ potential for delivering effective reward signals. In puzzle reasoning of LLMs, existing benchmarks evaluate different types of reasoning, including abstract, deductive, and compositional reasoning. Few benchmarks support scalable generation and difficulty control but lack puzzle diversity. Moreover, the improvement of LLMs’ puzzle-solving abilities mainly falls into two categories: tool integration and RLVR. Researchers from ByteDance Seed, Fudan University, Tsinghua University, Nanjing University, and Shanghai Jiao Tong University have proposed Enigmata, the first comprehensive toolkit designed for improving LLMs with puzzle reasoning skills. It contains 36 tasks across seven categories, each featuring a generator that produces unlimited examples with controllable difficulty and a rule-based verifier for automatic evaluation. The researchers further developed Enigmata-Eval as a rigorous benchmark and created optimized multi-task RLVR strategies. Puzzle data from Enigmata enhances SoTA performance on advanced math and STEM reasoning tasks like AIME, BeyondAIME, and GPQA when trained on larger models like Seed1.5-Thinking. This shows the generalization benefits of Enigmata. The Enigmata-Data comprises 36 puzzle tasks organized into 7 primary categories, including Crypto, Arithmetic, Logic, Grid, Graph, Search, and Sequential Puzzle, making it the only dataset having multiple task categories with scalability, automatic verification, and public availability. The data construction follows a three-phase pipeline: Tasks Collection and Design, Auto-Generator and Verifier Development, and Sliding Difficulty Control. Moreover, the Enigmata-Eval is developed by systematically sampling from the broader dataset, aiming to extract 50 instances per difficulty level for each task. The final evaluation set contains 4,758 puzzle instances rather than the theoretical maximum of 5,400, due to inherent constraints, where some tasks generate fewer instances per difficulty level. The proposed model outperforms most public models on Enigmata-Eval with 32B parameters, showing the effectiveness of the dataset and training recipe. The model stands out on the challenging ARC-AGI benchmark, surpassing strong reasoning models such as Gemini 2.5 Pro, o3-mini, and o1. The Qwen2.5-32B-Enigmata shows outstanding performance in structured reasoning categories, outperforming in Crypto, Arithmetic, and Logic tasks, suggesting effective development of rule-based reasoning capabilities. The model shows competitive performance in search tasks that require strategic exploration and planning capabilities. Moreover, Crypto and Arithmetic tasks tend to provide the highest accuracy, while spatial and sequential tasks remain more difficult. In this paper, researchers introduced Enigmata, a comprehensive suite for equipping LLMs with advanced puzzle reasoning that integrates seamlessly with RL using verifiable rule-based rewards. The trained Enigmata-Model shows superior performance and robust generalization skills through RLVR training. Experiments reveal that when applied to larger models such as Seed1.5-Thinking (20B/200B parameters), synthetic puzzle data brings additional benefits in other domains, including mathematics and STEM reasoning over state-of-the-art models. Enigmata provides a solid foundation for the research community to advance reasoning model development, offering a unified framework that effectively bridges logical puzzle-solving with broader reasoning capabilities in LLMs. Check out the Paper, GitHub Page and Project Page. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 95k+ ML SubReddit and Subscribe to our Newsletter. Sajjad AnsariSajjad Ansari is a final year undergraduate from IIT Kharagpur. As a Tech enthusiast, he delves into the practical applications of AI with a focus on understanding the impact of AI technologies and their real-world implications. He aims to articulate complex AI concepts in a clear and accessible manner.Sajjad Ansarihttps://www.marktechpost.com/author/sajjadansari/Multimodal Foundation Models Fall Short on Physical Reasoning: PHYX Benchmark Highlights Key Limitations in Visual and Symbolic IntegrationSajjad Ansarihttps://www.marktechpost.com/author/sajjadansari/Meta AI Introduces Multi-SpatialMLLM: A Multi-Frame Spatial Understanding with Multi-modal Large Language ModelsSajjad Ansarihttps://www.marktechpost.com/author/sajjadansari/Can LLMs Really Judge with Reasoning? Microsoft and Tsinghua Researchers Introduce Reward Reasoning Models to Dynamically Scale Test-Time Compute for Better AlignmentSajjad Ansarihttps://www.marktechpost.com/author/sajjadansari/NVIDIA AI Introduces AceReason-Nemotron for Advancing Math and Code Reasoning through Reinforcement Learning

Warning: contains spoilers for Doctor Who episode “The Reality War.” In Doctor Who’s frankly mind-boggling season finale, the Doctor’s epic battle with the two Ranis, Omega, Conrad and a herd of skyscraper-sized bone creatures ultimately comes down to the restoration of a single life – and will require a sacrifice nobody expected. Spoilers ahead. It’s honestly difficult to know where to start with this episode. There are so many potential jumping-off points for discussion – though, somewhat tellingly, very few of them relate to the actual story that kicked off in earnest last week, which the episode itself seems positively impatient to get out of the way. It takes about 15 minutes for the Doctor to stop hugging every member of the extended supporting cast so the titular war can kick off, then by the halfway mark it’s over. Audacious? Yes, though that’s not to say it actually works. Do we start with Billie Piper? Or the unexpected and quite charming Jodie Whittaker cameo? Or the fact that they somehow snuck Ncuti Gatwa’s regeneration onto the screen without anybody knowing? No, because this season didn’t start with Billie Piper, or Jodie Whittaker, or the Rani, or Ruby Sunday. It didn’t even start with the Doctor. It started with Belinda Chandra. A character with so much potential – compassionate, uncertain, a little bit spiky, competent in a new and interesting way, compellingly distrustful of the Doctor. Potential that has, at this point, been mostly wasted. There is a point in “The Reality War” where Belinda basically tells the Doctor “OK I think I’m done contributing to this episode, good luck tho” and is left holding the baby in a soundproofed box where she can neither affect or be affected by the story happening outside. We even have an unintentionally comical cut back to her standing in there, doing nothing, saying nothing. It’s hard to think of a more literal way to sideline a key player. This is the co-lead of the show! The companion! And instead of having any real agency, instead of contributing to the plot in any meaningful way whatsoever, she functionally stops existing as a narrative presence. She doesn’t even get to go with the Doctor when he rushes off to save what turns out to be her child. And for what? So that the companion who supposedly left the show last season can have all the big dramatic moments instead? There were no advanced screeners available for this episode – given what happens at the end, it’s easy to see what they were scared might leak – so I’m writing with less distance than usual, reacting fairly rapidly to a first watch. But even with several days to digest, it’s difficult to imagine feeling anything other than bafflement at this storytelling choice. This is what Belinda’s whole story arc was leading to? This is the big twist? It’s truly one of the most bewildering decisions that Russell T Davies has made. It already kind of felt like he’d run out of meaningful stuff for Belinda to do after “The Well”, and there have been plenty of complaints about her sidelining in “Wish World”, but nobody could have predicted this. Sorry Belinda. And sorry Varada Sethu. You both deserved better. Now to Ncuti Gatwa. It’s pointless getting into behind-the-scenes gossip, or speculating on the actor’s motivations – if he only ever wanted to do two seasons, of course that’s his choice. But what are we to take away from his brief tenure in storytelling and character terms? A Doctor defined by his joy, his exuberance, his love for people (frustratingly, a point this episode hammers until it becomes tedious). A smile as powerful as a billion supernovas. A killer wardrobe. Even in lesser episodes, Gatwa’s energy has carried us along, infectious and delightful. It’s a genuine shock and a shame to see him go. Join our mailing list Get the best of Den of Geek delivered right to your inbox! Not least because, as with Belinda, it feels like his Doctor had unfulfilled potential. Was this episode truly a satisfying conclusion to Fifteen’s story? He gets plenty of good moments, big and small, and of course he plays the hell out of all of them. You could argue that sacrificing their life to save one child is about the most Doctor-ish thing possible. I wouldn’t necessarily argue with you. But that’s broad strokes stuff, generally applicable to any incarnation. What about this Doctor makes this specific set of circumstances a fitting send-off? Is it satisfying for this Doctor, a Doctor representing a particular streak of joyful hedonism, a Doctor who releases UNIT from their stifling roles in Conrad’s reactionary wish world via an explicit and triumphant assertion of his queerness, to go out in this way, for these reasons? It just doesn’t feel like that’s what these past two seasons – the bi-generation, his relationship with Rogue, his torturing of Kid, the seemingly forgotten Susan stuff – have been leading to. It’s a shame that the episode also feels so messy on a minute-to-minute level. There are individually effective moments – Dark Souls boss Omega is a fantastic visual, and him casually munching The Rani is enjoyably WTF (though I can’t help wishing they’d offed the other one and kept Archie Panjabi around). The moment with the Doctor and Belinda passing Poppy’s jacket back and forth and folding it until it vanishes is kind of jaw-dropping in how understated and upsetting it is. Anita’s first joke about being in hospitality is funny (the second and third iterations not so much). Millie Gibson does a great job, even if it feels like a misstep to give Ruby so much heavy lifting to do instead of Belinda. But the whole thing feels so all over the place that not even Gatwa’s megastar energy can hold it together. And now he’s gone, regenerated into Billie Piper. At this point, we have no idea when the show will be back. It’s impossible to know where this is going. And it’s hard not to feel torn – on the one hand, Billie Piper is a fantastic actor, and it’s fascinating to consider what her take on the role will be (though it should be noted that the credits pointedly don’t say “Billie Piper as The Doctor”, whatever that could mean). On the other hand, didn’t we just do this? We had the second Tennant Doctor, it was a lovely gift for fans that wrapped up some loose ends and gave everyone a big warm glow for the anniversary, and then we flew off with Ncuti Gatwa, an actor who couldn’t have screamed more loudly that things were going to be different. But now we’re looking backwards again. And as fun a surprise as Piper’s appearance is, as fully as she will no doubt own the role… it feels like another retrograde move. It’s Doctor Who celebrating itself, getting lost in its own mythos, turning inward. And so we end this oh-so promising season in a strange, unsettling place. An episode that doesn’t really seem to care that much about the story it claimed to be telling, which makes discussing it seem weirdly beside the point. A show in limbo. A whole incarnation of the Doctor gone, when we’d barely started to get to know him. A promising companion wasted. A showrunner everyone expected to be a safe pair of hands making some utterly confounding choices. Where do we go from here? Doctor Who series 15 is available to stream now on BBC iPlayer in the UK and on Disney+ around the world.