The new company is betting that robots cannot learn fast enough, cheaply enough, or safely enough through physical trial and error—and that world models can move that work into simulation.

2 min

Aug 22, 2026Products

Launch Products

OpenAI’s ChatGPT Teens Pitches Guidance Over Answers. Teachers See a Harder Classroom Test.

Study Mode is built to steer students through questions rather than hand over homework answers. Whether that design improves learning, curbs cheating, or simply moves the problem elsewhere remains unsettled.

3 min

Aug 22, 2026Infrastructure

Partnership Infrastructure

Microsoft and Qcells Want AI Data Centers to Bring Their Own Power

The proposed model would pair Microsoft’s expanding data-center footprint with newly built generation, flexible batteries and utility-facing capacity. The key test is whether exploration turns into projects that can genuinely match the load they add.

3 min

Aug 22, 2026Products

Launch Products

Serval’s Catalyst Pitches AI-Built Workflows as a ServiceNow Replacement

The new agent can turn recurring ticket problems into proposed automations, but Serval must still show it can manage the customization, governance and trust demands of large enterprises.

3 min

Aug 22, 2026Models

Open release Models

Roblox Opens Three AI Safety Models for Child-Protection Teams

The models give other services tools to test and adapt, but Roblox’s own metrics and research on messages that slipped through moderation show automation remains an incomplete safeguard.

2 min

Aug 22, 2026Infrastructure

Enterprise adoption Infrastructure

Hudson River Trading Makes a Multibillion-Dollar Bet on CoreWeave Compute

The agreement gives a major quantitative trading firm earlier broad access to Nvidia’s Vera Rubin systems, while testing whether specialized AI cloud capacity can become a deeper financial-services business.

3 min

FROM OUR TEAM

Superpower ChatGPT

Make room for your best ideas.

Turn ChatGPT into an organized, searchable AI workspace.

Explore Superpower ChatGPT

Aug 22, 2026Models

Launch Models

Ox Alpha Offers a Million Tokens, but Not a Name

The anonymous model exposes multimodal and agent-oriented features, but its architecture, benchmark record and price after the free preview remain unknown.

2 min

Aug 22, 2026Infrastructure

Infrastructure

AI’s Data-Center Bet Faces a $2.5T-to-$10T Revenue Hurdle

The two calculations use different financial assumptions, but each points to a monetization challenge far larger than the AI revenue levels described today.

2 min

Aug 22, 2026Infrastructure

Launch Infrastructure

AWS Brings GPT-5.6 Cross-Region Inference, With a Data-Location Choice

The new Bedrock profiles let applications draw from a wider compute pool without changing their core model calls. The trade-off is explicit: global routing offers the broadest capacity, while US routing keeps processing within that geography.

3 min

Aug 22, 2026Models

Benchmark Models

Stanford Won Databricks’ Agent Cup, but 18.8% of Questions Stumped Every Team

The live contest tested whether agents developed on one benchmark could handle an unfamiliar Treasury archive. Stanford led the field, but the unsolved questions show the limits of reliable document reasoning.

2 min

Aug 22, 2026Products

Launch Products

Databricks Pushes Feature Stores From Batch Lag to 200ms Freshness

The company’s new streaming path targets decisions that change faster than scheduled data jobs can run. Its 200ms p99 figure is company-reported, and the operational trade-off is up to five minutes of replayed Kafka data after a failure.

3 min

Aug 22, 2026Products

Launch Products

Databricks Says Its New Extraction Mode Beats Frontier Models on the Documents That Break Them

Precision Mode is designed for cross-page reasoning, enormous line-item outputs and deeply nested schemas. Databricks reports a seven-point lead over its best tested chunk-and-merge baseline, but the evaluation blends internal and public datasets.

3 min

Aug 22, 2026Policy

Government action Policy

FDA Opens Door to Clinician-Style Tests for Medical AI

The agency is exploring a shift from attempting to test every possible AI interaction toward proving that a finished device can safely perform a defined clinical role—and continue doing so after changes.

3 min

Aug 22, 2026Models

Open release Models

Thinking Machines Releases Inkling, but Argues Open Weights Need a Gate

The lab’s argument is not that every capable model should be closed. It is that release decisions should turn on whether a model adds risk beyond what is already downloadable—and whether defenders have had time to prepare.

3 min

Aug 22, 2026Products

Launch Products

Grok Bot Gives Each Team Member a Persistent AI Workspace

The product gives each worker a shared cloud computer for multiple long-lived Bots, extending AI work into browsing, tools and local-machine actions while leaving several enterprise controls unfinished.

3 min

Aug 22, 2026Business

Enterprise adoption Business

Anthropic Leads Ramp’s July AI Buyer Panel, but OpenAI Is Gaining Ground

The completed July snapshot still favors Anthropic, while a faster OpenAI growth signal points to a contest that could change before the quarter closes.

3 min

Aug 22, 2026Products

Benchmark Products

Oracle Wants AI Agents to Pick Trusted Reports, Not Write SQL

The design can make database answers more repeatable by fixing the query underneath them. But it leaves a consequential judgment with the model: selecting the right report and inputs.

3 min

Aug 22, 2026Models

Benchmark Models

GLM-5.3’s Cheap Retries Put Fable 5’s Coding Premium Under Pressure

Together AI’s benchmark run suggests first-attempt parity is not enough to justify Fable 5’s price for general coding. But a Rust and serialization advantage, plus benchmark-specific cost assumptions, leave room for targeted routing.

2 min

Load more stories18 more next

Puzzle break

Daily games

Finding today’s puzzle for you.

New puzzles every day

Explore games

Puzzle break

Daily games

Finding today’s puzzle for you.

New puzzles every day

Explore games