Home

a shaneful place

News

Artificial Intelligence

Blog

Portfolio

Subscribe

AI News Digest - 2026-07-15

Category

Empty

1.

Anthropic found a hidden internal "J-space" in its Claude models that contained tokens not present in outputs but that appeared to influence how the models reasoned, and it published research describing the discovery and its implications for interpretability and monitoring.

2.

DeepMind CEO Demis Hassabis proposed creating a new U.S. standards body modeled on FINRA to develop evaluation protocols for frontier AI models and to coordinate possible slowdowns, calling for guardrails to manage advanced AI development.

3.

Google added AI image generation to Search's AI Overviews, enabling its Nano Banana 2 Lite model to generate images when no matching web images were found and beginning a staged rollout in the coming weeks.

4.

OpenAI re-enabled ChatGPT on WhatsApp across the European Economic Area after EU measures required Meta to open its platform to rival AI bots, restoring the service in the 27 EU member states plus Liechtenstein, Iceland, and Norway.

5.

Anthropic launched Claude for Teachers as a free offering for verified K‑12 educators in U.S. schools and stated that it would not train its models on student data.

References

1.

https://www.technologyreview.com/2026/07/13/1140343/what-anthropics-latest-ai-discovery-does-and-doesnt-show/

What Anthropic’s latest AI discovery does—and doesn’t—show

The company says it has found a new window into how its models arrive at answers. We spoke with senior editor Will Douglas Heaven about it.

technologyreview.com

1.

https://the-decoder.com/deepmind-ceo-hassabis-says-nobody-in-the-world-knows-what-happens-next-so-cautious-optimism-means-building-guardrails-now/

Deepmind CEO Hassabis says "nobody in the world knows what happens next" so "cautious optimism" means building guardrails now

Google Deepmind CEO Demis Hassabis has published a sweeping proposal for how to handle advanced AI. He wants a new US standards body modeled after financial regulator FINRA that would develop evaluation protocols for frontier models and could coordinate a slowdown in AI development if needed. Startups and research models would be exempt.

the-decoder.com

1.

https://the-decoder.com/google-search-now-generates-ai-images-when-it-cant-find-what-youre-looking-for-on-the-web/

Google Search now generates AI images when it can't find what you're looking for on the web

Google is adding AI image generation to Search's AI Overviews. When no matching image exists on the web, the new Nano Banana 2 Lite model generates one from the search query. The rollout starts in the coming weeks.

the-decoder.com

1.

https://the-decoder.com/chatgpt-returns-to-whatsapp-in-europe-after-eu-forces-meta-to-open-the-door-to-rival-ai-bots/

ChatGPT returns to WhatsApp in Europe after EU forces Meta to open the door to rival AI bots

OpenAI has re-enabled ChatGPT on WhatsApp, but only in the European Economic Area, covering the 27 EU member states plus Liechtenstein, Iceland, and Norway.

the-decoder.com

1.

https://the-decoder.com/anthropic-opens-claude-for-teachers-with-a-promise-not-to-train-models-on-student-data/

Anthropic opens Claude for Teachers with a promise not to train models on student data

Anthropic is rolling out Claude for Teachers, a free offering for verified K-12 educators at US schools.

the-decoder.com

Made with Slashpage