The Wacom One 14 is a computer-tethered pen display that tries to pull artists away from the iPad , but its solid specifications can't fend off a changing market forever. Wacom One 14 As a professional digital illustrator with 15+ years of experience across comics, gaming, and everything in between, I love pen displays. Pen display tablets and digital art are vital to my day-to-day workflow and productivity. My very first pen display was a Wacom Cintiq, and for many, many years used Wacom products exclusively for all of my illustration needs. Continue Reading on AppleInsider | Discuss on our Forums

• AppleInsider
• 4 gün, 22 saat, 14 dakika

Tyler Katzenberger / Politico : California Gov. Gavin Newsom vetoes SB 771, which would've fined social media companies if their algorithms intentionally promoted violent or extremist content —  Newsom, in a statement explaining his veto, said he shared concerns about “discriminatory threats, violence and coercive harassment online” but called SB771 “premature.”

• Techmeme
• 5 gün, 2 saat, 15 dakika

Michael Kan / PCMag : Nvidia says it will begin selling the DGX Spark mini PC for AI developers on October 15 on Nvidia.com and select third-party retailers for $3,999 —  (PCMag/Michael Kan) … It's not a consumer desktop, but Nvidia's foray into an AI developer-focused mini PC is finally ready to launch.

• Techmeme
• 5 gün, 2 saat, 25 dakika

Enhancements will make the simple bot act more like an AI chatbot.

• CNET
• 4 gün, 23 saat

Looking for help with today's New York Times Pips? We'll walk you through today's puzzle and help you match dominoes to tiles.

• Forbes Tech
• 5 gün, 3 saat, 6 dakika

Yoolim Lee / Bloomberg : Samsung projects Q3 operating profit up 32% YoY to ~$8.47B vs. ~$6.8B est., its biggest quarterly profit in more than three years, as AI development accelerates —  Samsung Electronics Co. posted its biggest quarterly profit in more than three years, reflecting booming memory chip demand while AI development accelerates globally.

• Techmeme
• 5 gün, 3 saat, 25 dakika

Andrew J. Hawkins / The Verge : Microsoft unveils MAI-Image-1, its first text-to-image AI model developed in house, and says it excels at photorealistic imagery, like lighting and landscapes —  The model has already secured a spot in the top 10 of LMArena. … Microsoft AI just announced its first text-to-image generator …

• Techmeme
• 5 gün, 3 saat, 30 dakika

Researchers at the Massachusetts Institute of Technology (MIT) are gaining renewed attention for developing and open sourcing a technique that allows large language models (LLMs) — like those underpinning ChatGPT and most modern AI chatbots — to improve themselves by generating synthetic data to fine-tune upon. The technique, known as SEAL (Self-Adapting LLMs), was first described in a paper published back in June and covered by VentureBeat at the time. A significantly expanded and updated version of the paper was released last month , as well as open source code posted on Github (under an MIT License, allowing for commercial and enterprise usage), and is making new waves among AI power users on the social network X this week. SEAL allows LLMs to autonomously generate and apply their own fine-tuning strategies. Unlike conventional models that rely on fixed external data and human-crafted optimization pipelines, SEAL enables models to evolve by producing their own synthetic training data and corresponding optimization directives. The development comes from a team affiliated with MIT’s Improbable AI Lab, including Adam Zweiger, Jyothish Pari, Han Guo, Ekin Akyürek, Yoon Kim, and Pulkit Agrawal. Their research was recently presented at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025). Background: From “Beyond Static AI” to Self-Adaptive Systems Earlier this year, VentureBeat first reported on SEAL as an early-stage framework that allowed language models to generate and train on their own synthetic data — a potential remedy for the stagnation of pretrained models once deployed. At that stage, SEAL was framed as a proof-of-concept that could let enterprise AI agents continuously learn in dynamic environments without manual retraining. Since then, the research has advanced considerably. The new version expands on the prior framework by demonstrating that SEAL’s self-adaptation ability scales with model size, integrates reinforcement learning more effectively to reduce catastrophic forgetting, and formalizes SEAL’s dual-loop structure (inner supervised fine-tuning and outer reinforcement optimization) for reproducibility. The updated paper also introduces evaluations across different prompting formats, improved stability during learning cycles, and a discussion of practical deployment challenges at inference time. Addressing the Limitations of Static Models While LLMs have demonstrated remarkable capabilities in text generation and understanding, their adaptation to new tasks or knowledge is often manual, brittle, or dependent on context. SEAL challenges this status quo by equipping models with the ability to generate what the authors call “self-edits” — natural language outputs that specify how the model should update its weights. These self-edits may take the form of reformulated information, logical implications, or tool configurations for augmentation and training. Once generated, the model fine-tunes itself based on these edits. The process is guided by reinforcement learning, where the reward signal comes from improved performance on a downstream task. The design mimics how human learners might rephrase or reorganize study materials to better internalize information. This restructuring of knowledge before assimilation serves as a key advantage over models that passively consume new data “as-is.” Performance Across Tasks SEAL has been tested across two main domains: knowledge incorporation and few-shot learning. In the knowledge incorporation setting, the researchers evaluated how well a model could internalize new factual content from passages similar to those in the SQuAD dataset, a benchmark reading comprehension dataset introduced by Stanford University in 2016, consisting of over 100,000 crowd-sourced question–answer pairs based on Wikipedia articles (Rajpurkar et al., 2016). Rather than fine-tuning directly on passage text, the model generated synthetic implications of the passage and then fine-tuned on them. After two rounds of reinforcement learning, the model improved question-answering accuracy from 33.5% to 47.0% on a no-context version of SQuAD — surpassing results obtained using synthetic data generated by GPT-4.1. In the few-shot learning setting, SEAL was evaluated using a subset of the ARC benchmark, where tasks require reasoning from only a few examples. Here, SEAL generated self-edits specifying data augmentations and hyperparameters. After reinforcement learning, the success rate in correctly solving held-out tasks jumped to 72.5%, up from 20% using self-edits generated without reinforcement learning. Models that relied solely on in-context learning without any adaptation scored 0%. Technical Framework SEAL operates using a two-loop structure: an inner loop performs supervised fine-tuning based on the self-edit, while an outer loop uses reinforcement learning to refine the policy that generates those self-edits. The reinforcement learning algorithm used is based on ReSTEM, which combines sampling with filtered behavior cloning. During training, only self-edits that lead to performance improvements are reinforced. This approach effectively teaches the model which kinds of edits are most beneficial for learning. For efficiency, SEAL applies LoRA-based fine-tuning rather than full parameter updates, enabling rapid experimentation and low-cost adaptation. Strengths and Limitations The researchers report that SEAL can produce high-utility training data with minimal supervision, outperforming even large external models like GPT-4.1 in specific tasks. They also demonstrate that SEAL generalizes beyond its original setup: it continues to perform well when scaling from single-pass updates to multi-document continued pretraining scenarios. However, the framework is not without limitations. One issue is catastrophic forgetting, where updates to incorporate new information can degrade performance on previously learned tasks. In response to this concern, co-author Jyo Pari told VentureBeat via email that reinforcement learning (RL) appears to mitigate forgetting more effectively than standard supervised fine-tuning (SFT), citing a recent paper on the topic. He added that combining this insight with SEAL could lead to new variants where SEAL learns not just training data, but reward functions. Another challenge is computational overhead: evaluating each self-edit requires fine-tuning and performance testing, which can take 30–45 seconds per edit — significantly more than standard reinforcement learning tasks. As Jyo explained, “Training SEAL is non-trivial because it requires 2 loops of optimization, an outer RL one and an inner SFT one. At inference time, updating model weights will also require new systems infrastructure.” He emphasized the need for future research into deployment systems as a critical path to making SEAL practical. Additionally, SEAL’s current design assumes the presence of paired tasks and reference answers for every context, limiting its direct applicability to unlabeled corpora. However, Jyo clarified that as long as there is a downstream task with a computable reward, SEAL can be trained to adapt accordingly—even in safety-critical domains. In principle, a SEAL-trained model could learn to avoid training on harmful or malicious inputs if guided by the appropriate reward signal. AI Community Reactions The AI research and builder community has reacted with a mix of excitement and speculation to the SEAL paper. On X, formerly Twitter, several prominent AI-focused accounts weighed in on the potential impact. User @VraserX , a self-described educator and AI enthusiast, called SEAL “the birth of continuous self-learning AI” and predicted that models like OpenAI's GPT-6 could adopt similar architecture. In their words, SEAL represents “the end of the frozen-weights era,” ushering in systems that evolve as the world around them changes. They highlighted SEAL's ability to form persistent memories, repair knowledge, and learn from real-time data, comparing it to a foundational step toward models that don’t just use information but absorb it. Meanwhile, @alex_prompter , co-founder of an AI-powered marketing venture, framed SEAL as a leap toward models that literally rewrite themselves. “MIT just built an AI that can rewrite its own code to get smarter,” he wrote. Citing the paper’s key results — a 40% boost in factual recall and outperforming GPT-4.1 using self-generated data — he described the findings as confirmation that “LLMs that finetune themselves are no longer sci-fi.” The enthusiasm reflects a broader appetite in the AI space for models that can evolve without constant retraining or human oversight — particularly in rapidly changing domains or personalized use cases. Future Directions and Open Questions In response to questions about scaling SEAL to larger models and tasks, Jyo pointed to experiments (Appendix B.7) showing that as model size increases, so does their self-adaptation ability. He compared this to students improving their study techniques over time — larger models are simply better at generating useful self-edits. When asked whether SEAL generalizes to new prompting styles, he confirmed it does, citing Table 10 in the paper. However, he also acknowledged that the team has not yet tested SEAL’s ability to transfer across entirely new domains or model architectures. “SEAL is an initial work showcasing the possibilities,” he said. “But it requires much more testing.” He added that generalization may improve as SEAL is trained on a broader distribution of tasks. Interestingly, the team found that only a few reinforcement learning steps already led to measurable performance gains. “This is exciting,” Jyo noted, “because it means that with more compute, we could hopefully get even more improvements.” He suggested future experiments could explore more advanced reinforcement learning methods beyond ReSTEM, such as Group Relative Policy Optimization (GRPO). Toward More Adaptive and Agentic Models SEAL represents a step toward models that can autonomously improve over time, both by integrating new knowledge and by reconfiguring how they learn. The authors envision future extensions where SEAL could assist in self-pretraining, continual learning, and the development of agentic systems — models that interact with evolving environments and adapt incrementally. In such settings, a model could use SEAL to synthesize weight updates after each interaction, gradually internalizing behaviors or insights. This could reduce the need for repeated supervision and manual intervention, particularly in data-constrained or specialized domains. As public web text becomes saturated and further scaling of LLMs becomes bottlenecked by data availability, self-directed approaches like SEAL could play a critical role in pushing the boundaries of what LLMs can achieve. You can access the SEAL project, including code and further documentation, at: https://jyopari.github.io/posts/seal

• VentureBeat
• 4 gün, 23 saat, 39 dakika

Microsoft is continuing to roll out in-house AI models, further decreasing its reliance on long-standing partnership with OpenAI. Today, the company introduced MAI-Image-1, its first internally-developed image-generating AI model. According to the blog post , MAI-Image-1 is particularly good for creating photorealistic results, and can generate natural lighting and landscapes. For now, the model is being tested on LMArena, and Microsoft said it plans to roll out MAI-Image-1 to Copilot and its Bing Image Creator "very soon." Over the summer, Microsoft made its first break from collaborating with OpenAI when it unveiled its first two in-house trained models, MAI-Voice-1 and MAI-1-preview. At that time, Microsoft AI division leader Mustafa Suleyman said in an interview that the company had "an enormous five-year roadmap that we're investing in quarter after quarter." So far, it's at least setting a solid clip of releases. This article originally appeared on Engadget at https://www.engadget.com/microsoft-debuts-its-first-in-house-ai-image-generator-224153867.html?src=rss

• Engadget
• 4 gün, 23 saat, 48 dakika

Enterprises often find that when they fine-tune models , one effective approach to making a large language model (LLM) fit for purpose and grounded in data is to have the model lose some of its abilities. After fine-tuning, some models “forget” how to perform certain tasks or other tasks they already learned. Research from the University of Illinois Urbana-Champaign proposes a new method for retraining models that avoids “catastrophic forgetting,” in which the model loses some of its prior knowledge. The paper focuses on two specific LLMs that generate responses from images: LLaVA and Qwen 2.5-VL. The approach encourages enterprises to retrain only narrow parts of an LLM to avoid retraining the entire model and incurring a significant increase in compute costs. The team claims that catastrophic forgetting isn’t true memory loss, but rather a side effect of bias drift. “Training a new LMM can cost millions of dollars, weeks of time, and emit hundreds of tons of CO2, so finding ways to more efficiently and effectively update existing models is a pressing concern,” the team wrote in the paper . “Guided by this result, we explore tuning recipes that preserve learning while limiting output shift.” The researchers focused on a multi-layer perceptron (MLP), the model's internal decision-making component. Catastrophic forgetting The researchers wanted first to verify the existence and the cause of catastrophic forgetting in models. To do this, they created a set of target tasks for the models to complete. The models were then fine-tuned and evaluated to determine whether they led to substantial forgetting. But as the process went on, the researchers found that the models were recovering some of their abilities. “We also noticed a surprising result, that the model performance would drop significantly in held out benchmarks after training on the counting task, it would mostly recover on PathVQA, another specialized task that is not well represented in the benchmarks,” they said. “Meanwhile, while performing the forgetting mitigation experiments, we also tried separately tuning only the self-attention projection (SA Proj) or MLP layers, motivated by the finding that tuning only the LLM was generally better than tuning the full model. This led to another very surprising result – that tuning only self-attention projection layers led to very good learning of the target tasks with no drop in performance in held out tasks, even after training all five target tasks in a sequence.” The researchers said they believe that “what looks like forgetting or interference after fine-tuning on a narrow target task is actually bias in the output distribution due to the task distribution shift.” Narrow retraining That finding turned out to be the key to the experiment. The researchers noted that tuning the MLP increases the likelihood of “outputting numeric tokens and a highly correlated drop in held out task accuracy.” What it showed is that a model forgetting some of its knowledge is only temporary and not a long-term matter. “To avoid biasing the output distribution, we tune the MLP up/gating projections while keeping the down projection frozen, and find that it achieves similar learning to full MLP tuning with little forgetting,” the researchers said. This allows for a more straightforward and more reproducible method for fine-tuning a model. By focusing on a narrow segment of the model, rather than a wholesale retraining, enterprises can cut compute costs. It also allows better control of output drift. However, the research focuses only on two models, specifically those dealing with vision and language. The researchers noted that due to limited resources, they are unable to try the experiment with other models. Their findings, however, can be extended to other LLMs, especially for different modalities.

• VentureBeat
• 4 gün, 23 saat, 51 dakika

The Information : Sources: OpenAI is working with Arm to develop a CPU designed to work with the AI chip OpenAI is developing with Broadcom; TSMC will manufacture the AI chip —  OpenAI's development of its own artificial intelligence chip will benefit SoftBank, one of its biggest shareholders …

• Techmeme
• 5 gün, 4 saat

Google is changing the game with smartwatch reparability, and Apple should take note.

• TechRadar
• 5 gün

Amazon has like-new Kindle Paperwhites on sale for just $96.99, which is about 28% less than buying a new model.

• Mashable
• 5 gün

Whether you’re looking for a productivity desktop, a gaming PC powerhouse, or a stylish all-in-one Windows machine, we’ve got you covered. The team at PCWorld sort through all of the daily computer sales and put together a curated list of the best deals available. But not all deals are really deals, so we only choose those offered by reputable companies and that include great hardware to ensure you get the best value for your money. We’ve also included some helpful answers to common questions about buying a computer at the bottom of this article. If you’re considering a laptop instead, be sure to check out our best laptop deals , updated daily. Note: Tech deals come and go quickly, so it’s possible some of these computer discounts will have expired before this article’s next update. Best gaming desktop computer deals Skytech Storm , Ryzen 7 5700/RTX 5060 Ti/16GB RAM/1TB SSD, $999.99 (12% off on Amazon) Acer Nitro 60 , Ryzen 7 7700/RTX 5070/32GB RAM/2TB SSD, $1,369.00 (28% off on Amazon) LXZ Gaming PC , Ryzen 7 8700F/RX 7650 GRE/32GB RAM/1TB SSD, $899.99 (10% off on Amazon) Skytech Azure , Ryzen 7 5700/RTX 5060/32GB RAM/1TB SSD, $999.99 (12% off on Amazon) iBuyPower Y40 , Ryzen 9 7900X/RTX 5070 Ti/32GB RAM/2TB SSD, $2,069.99 (10% off on Amazon) Alienware Aurora , Core Ultra 9 285K/RTX 5080/32GB RAM/2TB SSD, $2,549.99 (20% off on Dell) HP Omen 35L , Core Ultra 5 225F/RTX 5060/16GB RAM/512GB SSD, $1,119.99 (30% off on HP) My top picks: The Skytech Storm for $130 off on Amazon is the budget deal of this week. With a Ryzen 7 5700 CPU and RTX 5060 Ti it’ll deliver solid frame rates at 1080p for modern games and the inclusion of 1TB of onboard storage means you’ll have plenty of space to load it up with your game library. Even though October Prime Day wrapped up, Amazon still continues to have excellent deals on gaming PCs. The Acer Nitro 60 for $530 off is an outrageous price for a strong midrange rig like this. Not only do you get an RTX 5070, but it also comes with 32GB of RAM and a generous 2TB of onboard SSD storage. Best mini-PC deals GMKtec M7 , Ryzen 7 6850H/32GB RAM/512GB SSD, $365.98 (25% off on Amazon) GMKtec K12 , Ryzen 7 H 255/Radeon 780M/32GB RAM/512GB SSD, $499.99 (31% off on Amazon) AceMagician K1 Mini PC , Ryzen 7 5700U/16GB RAM/512GB SSD, $279.00 (30% off on Amazon) AceMagic Vista Mini N1 , Alder Lake-N N97/16GB RAM/512GB SSD, $179.00 (31% off on Amazon) KAMRUI E3B , Ryzen 5 7430 U/32GB RAM/512GB SSD, $299.99 (25% off on Amazon) My top picks: Amazon is offering the GMKtec M7 mini-PC for $134 off. This mini-PC not only rocks a Ryzen 7 Pro 6850H CPU and a whopping 32GB of RAM, but comes with excellent connectivity features to boot. It’s a great mini-PC and at this discount the value can’t be beat. Another GMKtec deal on Amazon is also a highlight deal right now. The GMKtec K12 for $220 packs a solid Ryzen 7 CPU and a Radeon 780M GPU—graphics cards are a rarity in mini-PCs—so you can even do some gaming on this bad boy. Best all-in-one computer deals All-in-one desktop computers combine a PC’s hardware with a modern display to make a desktop computer that has both form and function. Since everything is built together, you can save precious desktop space with an all-in-one. They make capable work computers and they can also be excellent home computers with the wide range of features appealing to the whole family. Lenovo 24 AiO , Core i5-1140G7/32GB RAM/1TB SSD/24-inch 1080p display, $599.99 (33% off on Amazon) HP OmniStudio X AiO , Core Ultra 5 125H/16GB RAM/256GB SSD/32-inch 4K display, $1,189.99 (22% off on HP) HP 24 AiO , Core i3-1110G4/32GB RAM/1TB SSD/24-inch 1080p touch display, $699.99 (63% off on Amazon) Dell 27 AiO , Core 5 120U/16GB RAM/512GB SSD/27-inch 1080p display, $879.99 (15% off on Dell) iMac M4 , M4/16GB RAM/512GB SSD/24-inch 5K display, $1,523.86 (10% off on Amazon) My top picks: Amazon is offering a great deal of $300 off on the Lenovo 24 AiO . This budget-friendly all-in-one has way more RAM and onboard storage than most other models at this price. It’s a killer value for a trustworthy and dependable Lenovo PC. Alternatively, you can opt for the HP 24 AiO for just $100 more and get practically the same computer with the addition of a touch screen instead. Alternatively, the HP OmniStudio X AiO for $340 off on HP’s website is a worthy splurge. Rocking a stunning 4K display and good performance features, this flagship all-in-one from HP will upgrade any desk space—you just might want to invest in some external storage as this only comes with 256GB. Computer deals FAQ 1. What are good websites to find computer deals? There are a ton of sites that sell computers, and scouring through all of them would take you a lot of time—that’s why we do it for you here and highlight the best deals we find. However, to save you some time and frustration, you need to be smart about where you look at any given time of the year. If you’re looking for a new computer during the holidays or around popular sale periods such as Black Friday or back-to-school, then you are likely to find great deals directly through first party vendor websites. These include the retail storefronts of popular computer manufacturers such as HP, Dell, and Lenovo. However, if you are looking in between sales periods, it’s generally a good idea to search through large third-party retailers such as Amazon , Adorama , Walmart , BestBuy , and Newegg . Oftentimes these websites will offer limited Deals of the Day type sales in hopes of getting rid of excess stock. On the upside, you can score still-decent PCs at a steep discount. 2. When’s the best time to shop for a PC computer? Typically you’ll want to time your PC computer shopping around a prominent sales period. The biggest sales periods are Black Friday/Cyber Monday in late November and Amazon Prime Day in early-to-mid July. The best sales often occur leading up to and during these two events and they are great times to snag a new PC computer for cheap. Other holiday shopping periods such as the New Year sales in January, President’s Day sales in April, and the back-to-school sales event in August are also good times to find discounts on computers. 3. What type of desktop should I get? You’ll see a ton of options when searching for a desktop computer, but they all mainly fit into four main categories: productivity tower PCs, gaming PCs, mini PCS, and all-in-ones (AiO). Which you should end up buying is entirely dependent upon what your needs are and what you want to do with your computer. If you are looking for something that will work in a home office or family room, then a productivity PC or AiO with a solid CPU and lots of RAM and storage is probably the way to go. If gaming is your main concern, a gaming PC can offer a lot more bang for your buck than a laptop and you should focus on getting the best GPU possible. Or if you just want something that can fit anywhere and provide basic computing then a mini PC is a good bet. 4. What CPU and GPU should I get? When looking at your new computer’s CPU, get at least an Intel Core i5 or AMD Ryzen 5, both of which will provide plenty of processing power for everyday computing tasks. If you don’t intend to do any PC gaming, then feel free to save some money by going with integrated graphics. However, if you are looking to get your game on, we recommend at least an Nvidia GeForce 3060 or AMD Radeon RX 6600 XT, as these are the least expensive discrete graphics cards that can handle ray tracing well. If you aren’t interested in those cutting-edge lighting effects, however, the RTX 3050 and Radeon RX 6600 also provide good 1080p gaming performance at even lower prices. 5. How much memory and storage does my PC need? As for RAM, we think its best to shoot for 16GB at the minimum for productivity and gaming, but for family computers and internet browsing, 8GB should suffice. Storage size is dependent upon your personal needs, but it is generally a good idea to opt for an SSD over an standard HDD as they are much faster and don’t significantly affect the price of a desktop. Before deciding, it’s best to consider what your intended use of the computer will be. Are you just doing work or web browsing? Then something like 512GB will be plenty. If you want to load up a lot of large files such as games or content creation projects, then you’ll need at least 1 or 2TB or storage. However, just remember that even if your computer doesn’t have enough storage built-in you can always upgrade your SSD or go with an external drive to increase your available storage options. 6. Is it a good idea to buy a refurbished computer? Refurbished computers are used machines that have been repaired, upgraded, and cleaned for the purpose of reselling. They’re usually open-box returns, overstock, or models with minor cosmetic damage (scratches, scuffs, etc). Refurbished computers can be a bargain hunter’s dream as they’re likely still in good (or great) condition and you can save a lot of money. That being said, refurbished computers can have their downsides as well. In addition to cosmetic blemishes, some of the internal components might be a little older or outdated and they might not be in peak condition due to previous usage. If you do consider buying a refurbished computer I recommend looking at eBay as they offer a one-year warranty. You can also check out manufacturer’s retail storefronts like Dell’s Outlet Store and Apple’s Refurbished Store —just be sure to look at the terms of warranty offered before purchasing.

• PCWorld
• 5 gün, 3 dakika

Ashley Capoot / CNBC : Charlie Kawwas, president of Broadcom's semiconductor group, says OpenAI is not the $10B customer the company announced during its earnings call in September —  Charlie Kawwas, president of the semiconductor solutions group at Broadcom, on Monday said that OpenAI is not the mystery $10 billion customer …

• Techmeme
• 5 gün, 4 saat, 10 dakika

We're live in San Francisco ahead of Dreamforce 2025.

• TechRadar
• 5 gün, 12 dakika