News Archive
Pick a date to read that day's stories.
Days with a blue dot have stories.
- 01 of 10ANTHROPIC
US Lifts Export Ban on Anthropic's Fable and Mythos Models
The US government removed export limits on Anthropic's Claude models. This allows the company to again offer these models to international users.
OUR TAKENothing like a national security panic to remind AI firms that governments still exist.
Read more → - 02 of 10OPENAI
OpenAI previews GPT-5.6 Sol, its next flagship model
OpenAI is testing its newest large computer model with partners. It will be cheaper than similar systems from competitors.
OUR TAKEHalf the price, available to twenty people, democracy in action.
Read more → - 03 of 10GOOGLE
Google ships Gemini Omni Flash for video generation
Google introduced a new video creation tool. It helps make videos and lets developers use it now.
OUR TAKE'Nano Banana 2 Lite' shipped in 2026 with no one apparently asking whether it should.
Read more → - 04 of 10NVIDIA
Etched exits stealth with $800M raised, $1B in orders
A new computer chip company, Etched, has secured $800 million and over $1 billion in orders for its new chips, aiming to challenge current market leaders.
OUR TAKEA billion in sales and no product page until Tuesday, stealth mode working overtime.
Read more → - 05 of 10OPENAI
OpenAI and Broadcom unveil first custom inference chip
OpenAI created a new chip with Broadcom, designed for their AI programs. It performs much better than current chips and helps OpenAI build its own supply chain.
OUR TAKENine months concept to chip, half designed by OpenAI's models, impressive, or slightly concerning.
Read more → - 06 of 10EU
EU agrees to delay key AI Act high-risk rules
EU lawmakers agreed to delay strict new rules for dangerous AI systems. The compliance deadline is now set for December 2027.
OUR TAKEBrussels grants another eighteen months on rules debated since 2021, urgency, EU-style.
Read more → - 07 of 10SAFETY
China's AI virtual companion rules take effect this month
New Chinese rules make sure websites offering AI friends must tell people they are not real. These rules also aim to stop users from getting too attached.
OUR TAKEChina has legally banned falling for chatbots, which rather suggests the problem is widespread.
Read more → - 08 of 10HEALTH
AI spots hidden cardiac arrest risk in routine ECGs
Researchers found a computer program can spot a hidden pattern in heart tests. This pattern shows who might be at higher risk of sudden cardiac arrest. This could improve routine heart checks.
OUR TAKEA century of ECG data, and it took a model to spot what cardiologists missed.
Read more → - 09 of 10SECURITY
Straiker raises $64M to secure the AI agent era
Straiker got $64 million to create security specifically for AI computer programs. Current security tools aren't good enough for these new, independent programs.
OUR TAKEWe built autonomous agents before the brakes, a sequencing choice now costing the industry dearly.
Read more → - 10 of 10FINANCE
MDOTM raises $27M as AI enters asset management
An investment firm got $27 million to help money managers use smart computer programs. This helps them make investment choices more efficiently.
OUR TAKEFund managers have spent decades insisting their edge is irreplaceable human intuition, and are now quietly funding its replacement.
Read more →