RESEARCH - AI Dynamics

AI accelerating scientific progress and ushering golden age

By

–

11 June 2026 21h05

My conversation with @ymatias (Head of Google Research) about how AI is accelerating the magic cycle of scientific progress, improving the lives of real people around the world, and us entering the golden age of research.

This chat left me feeling genuinely inspired : ) pic.twitter.com/SqN120OjFb
— Logan Kilpatrick (@OfficialLoganK) 11 juin 2026

My conversation with @ymatias (Head of Google Research) about how AI is accelerating the magic cycle of scientific progress, improving the lives of real people around the world, and us entering the golden age of research. This chat left me feeling genuinely inspired : )

→ View original post on X — @officiallogank

11 June 2026

DARPA AI Cyber Challenge: multi-agent LLM discovers zero-day vulnerabilities

By

@fogoros

–

11 June 2026 20h30

At the DARPA AI Cyber Challenge, Team Atlanta from Georgia Institute of Technology demonstrated a multi-agent LLM framework that discovered zero-day vulnerabilities in large codebases and automatically generated, tested, and deployed functional patches without human intervention.… pic.twitter.com/gBxsiw7Jgl
— Lucian Fogoros (@fogoros) 11 juin 2026

At the DARPA AI Cyber Challenge, Team Atlanta from Georgia Institute of Technology demonstrated a multi-agent LLM framework that discovered zero-day vulnerabilities in large codebases and automatically generated, tested, and deployed functional patches without human intervention.

→ View original post on X — @fogoros

11 June 2026

Perplexity Computer adds Deep Research as native skill

By

@aravsrinivas

–

11 June 2026 20h26

Perplexity Computer is an agent harness that just keeps delivering. Deep Research is now a native skill inside Computer (you don’t have to explicitly think of using it as a standalone mode anymore as long as you’re using Computer), furthering the state of the art significantly

→ View original post on X — @aravsrinivas

11 June 2026

Flywheel of models and data limited to verifiable domains

By

@jeande_d

–

11 June 2026 20h10

Models, synthetic data, and environments are a flywheel now. The best model generates the best data, trains a better model, generates even better data. Rinse and repeat. But the flywheel only spins where you can verify: math, code, agents with graders. Everywhere else it merely

→ View original post on X — @jeande_d

11 June 2026

Agents fail exam with Fable 5 included

By

@snorkelai

–

11 June 2026 20h07

New from Agents' Last Exam: every frontier agent tested, including Fable 5 (scored 0% on the benchmark's hardest tier of real professional work).

→ View original post on X — @snorkelai

11 June 2026

GPT-2 fears reminiscent of boy who cried wolf

By

@garymarcus

–

11 June 2026 19h48

they were terrified of GPT-2. definitely into boy who cried wolf territory.

→ View original post on X — @garymarcus

11 June 2026

Deep Research integrated as a native capability in Computer

By

@perplexity_ai

–

11 June 2026 19h32

We're integrating Deep Research as a native skill inside Computer.

It now connects to the agent harness that powers Computer, with access to search as code generation, long running sandboxes, connectors, tools, and licensed data.

Available now to Pro and Max subscribers. pic.twitter.com/uHpVISkh2P
— Perplexity (@perplexity_ai) 11 juin 2026

We are integrating Deep Research as a native capability inside Computer. It now connects to the agent harness that powers Computer, with access to research in the form of code generation, long-lived sandboxes, connectors,

→ View original post on X — @perplexity_ai

11 June 2026

Stanford HAI and AI4MH explore responsible AI for youth mental health

By

@stanfordhai

–

11 June 2026 19h17

Last week, @StanfordHAI & AI4MH's 2026 symposium explored responsible AI adoption in mental health. CA Assembly Member @AsmMiaBonta stressed the need to protect youth as they turn to chatbots & other AI tools for mental health challenges. Full session https://
youtube.com/watch?v=KBZJYk
zqT0c
…

→ View original post on X — @stanfordhai

11 June 2026

Gemini Omni Flash: SOTA in image, text to video and editing

By

@officiallogank

–

11 June 2026 19h05

Gemini Omni Flash is SOTA at image to video, text to video, and video editing : ) Excited to get this to developers in the API soon!

→ View original post on X — @officiallogank

11 June 2026

Use AI to think, not to skip thinking, says MIT Media Lab

By

@aihighlight

–

11 June 2026 19h01

9/ The fix is not to stop using AI. It is to make it explain how it got there, and to decide for yourself before you look. Use it to think. Not to skip thinking. Source: MIT Media Lab, presented at CHI 2026.

→ View original post on X — @aihighlight

11 June 2026