Baseten released bison-1 and groundhog-1, open-weight decision models built on Gemma 4. bison-1 scores 76 on the AA Decision Index v2.1, and both models support 128k context and image input.
With the watchOS 27.3 beta, Apple started testing Siri Recap. ‌Siri‌ Recap lets the Series 12 and Ultra 4 Apple Watch models listen to conversations to generate a brief summary of what was discussed. Actual audio is sectioned off from apps, Apple, and the rest of the OS, and then discarded after a summary is generated. It's a privacy-forward way of tracking what's going on around you, but after two weeks of testing, it's clear that the privacy safeguards limit its…
Senior Research Scientist Jim Glass discusses how his MIT-IBM Computing Research Lab collaboration connects perception and language, and addresses the shortcoming of current models.
OpenAI has significantly strengthened safeguards against extracting internal reasoning from GPT-6.2 Astra and GPT-6.2 Sol. In adaptive red-teaming tests, the success rate for extracting protected reasoning fell from 54 percent to 0.2 percent. The company is also suspending thousands to tens of thousands of accounts each week over coordinated distillation campaigns. The article OpenAI makes it much harder for attackers to extract how its latest AI models think appeared first…
An investigation led by the House Homeland Security Committee focused on the dangers of U.S. companies using lower-cost open weight AI models for certain projects.
Enforcing physics-based constraints prevents AI models from generating unreliable predictions for complex nonlinear systems in weather, climate, and turbulent flows.
The holiday shopping season is here, and Apple’s AirPods are sure to be one of the most popular gifts. If you’re shopping for AirPods, Apple sells three different models: AirPods 5, AirPods Pro 3, and AirPods Max 2. Here’s how they compare. Buy AirPods Pro 3 $179 (Reg. $249) Buy AirPods Max $419 (Reg. $549) Buy AirPods 5 from $129
Day 2 of of Pixel 11 Pro, Pixel 11 Pro XL, and Pixel 11 Pro, $100 off all Pixel Watch 5 models, the ongoing all-time low Google Home Speaker, and the official Nothing Black Friday deals at up to 49% off.
A new report from the AI evaluation nonprofit METR finds that some AI capabilities are continuing to improve along an exponential trajectory—but it also indicates that models aren’t quite as good at tasks that require judgment as they are at coding. The report is an update to the organization’s time horizon plot, which was widely…
Agentic AI is reviving an argument over renting versus owning compute, and the model plane — the layer that decides where AI models run — sits at the center of it. Open-source efforts such as Upbound Inc.’s Modelplane route inference across clouds, but many enterprises want the models on hardware they control. That shift favors
Day 2 of the giant Pixel 11 Pro, Pixel 11 Pro XL, and Pixel 11 Pro Fold as well as $100 off all Pixel Watch 5 models. These deals now join the lowest price we have ever tracked on the 2026 Moto Tag 2, a range of TCL NXTPAPER Android tabs from $190, and a chance to land a Galaxy Watch from just $200 over at Amazon. Just be sure to scope out the all-time low on Amazon’s all-new touchscreen Echo Time smart speaker and home hub, this giant collection of Nothing all-time lows…
Handing Michael Woodford and Robert Hall the Nobel Prize for Economics, the awarding committee said their work had helped deal with shocks. In truth, their models fed the illusion of stability that blinded policymakers to the looming 2008 crisis.
Nvidia is close to acquiring AI startup Inception for at least $500 million, according to The Information. The company is just two years old and builds language models based on diffusion processes, an alternative approach to the token-by-token generation used by models like ChatGPT. The deal would bolster Nvidia's own open-source models and its push into robotics. The article Nvidia bets at least $500 million on Inception's diffusion models to strengthen its open-source…
New images of dummy models offer an early look at the purported designs of the iPhone Air 2 and iPhone 18, showing a new dual rear camera arrangement for the Air and a largely unchanged appearance for the standard model. In a new post on X, leaker Sonny Dickson highlights the ‌iPhone Air‌'s move from one rear camera to two as the biggest change. The standard ‌iPhone 18‌, he says, retains a design similar to the iPhone 17. A second image shows the fronts…
Quantum 10, 2279 (2026). https://doi.org/10.22331/q-2026-11-20-2279 Motzkin spin-chains, which include `colorless' (integer spin $s=1$) and `colorful' ($s \geq 2$) variants, are one-dimensional (1D) local integer spin models notable for their lack of a conformal field theory (CFT) description of their low-energy physics, despite being gapless. The colorful variants are particularly unusual, as they exhibit power-law violation of the area-law of entanglement entropy (as…
Frontier models can exploit software flaws faster than the world can fix them. And the risks extend far beyond the cyber realm You’re not crazy to feel the world is coming apart at the seams. Within the last week, outages took down core services at Microsoft, United Airlines and JPMorgan Chase. On Friday, all US domestic departures were grounded for the second time since 9/11. And on Tuesday, CNN reported the NSA’s highly classified internal communications network went down…
This second installment reviews 13 more Amazon SageMaker AI inference launches from 2026, spanning configuration recommendations, fitting more models onto each GPU, serving models too large for one instance, cross-Region routing, and post-launch tuning, bringing the year's total to 26.
Kimi Code CLI v2.7.0 adds a native Amazon Bedrock provider, so you can connect the open source coding agent to models like Kimi K3.1 Pro using your standard AWS credential chain (named profiles or SSO) instead of a separate API key. This post shows how to configure the provider and send your first request.
On the day Anthropic went public, Fireworks AI’s co-founder made the case for open-source models as a cheaper alternative, and addressed the breaches that have cost the company some customers’ trust.
MedChemBench evaluates how well AI models interpret molecular structures, reason from experimental evidence, and select compounds in early drug discovery.
An MIT and Dana-Farber Institute-led team used large language models to examine nearly 4,000 studies and hundreds of code repositories, revealing gaps in the information needed to scrutinize and reuse published research.
Boltz PBC, a startup that develops artificial intelligence models for pharmaceutical researchers, has raised $110 million in funding. Bessemer led the Series A investment. Boltz stated on Wednesday that the fund was joined by a large group of new and returning backers. The consortium included Greycroft, Andreessen Horowitz, Amplify Partners, F-Prime Capital, Hummingbird and others.