Videos
Tutorials, deep dives, devlogs, and commentary on AI and agentic coding — from @UltraDotDev.
Sam Altman Gets To Meet Xi Xinping & Dario Amodei Isn't Invited #ai #openai #anthriopic #aislowdown
Sam Altman is attending the September 24th state dinner during the US-China summit. Anthropic isn't invited (yet). What is happening and what should we…
Anthropic Retains Accenture: Real Slowdown or Regulatory Capture #ai #anthropic #darioamodei #ainews
Anthropic has hired Accenture to help evaluate models and red-team. But this doesn't mean anything has changed...yet. We need real findings and real…
Jev Is A Crazy Powerful AI For...Making Decisions? #jev #ai #aicoding # #claudecode #codex
Jev is a new AI that works totally differently than LLMs. Instead of predicting text, it predicts events. People have used it to fly a drone in a…
OpenAI's GPT6 Astra Wrote Its Own Jailbreak? #ai #gpt6astra #codex #gpt6 #astra #openai #aisecurity
@OpenAI caught an unreleased gpt6 astra model writing jailbreaks for itself, starting while it was looking up...library books?
What's Hiding In Your AI Coding Junk Drawer? #ai #aicoding #claudecode #codex #fable5 #gpt6
That open-source tool you found on Product Hunt last night? 45 stars in a day isn't a background check.
Hey Claude, AI Is Coming For Your Job Too... #claude #kimiai #moonshotai #ai #claudecode
Anthropic says MoonshotAI (the company behind Kimi) routed customer questions to Claude and saved some exchanges to train its own models. Claude, buddy.…
Anthropic and OpenAI want to slow down. But will the US let them? #ai #anthropic #openai #agi
We're seeing more warnings than ever before about the dangers of agi and the singularity but are the warnings real or just anti-competitive? And if…
Is An AI Slowdown Possible? Or are we heading into an AI Cold War? #ai #anthropic #openai #aisafety
Are we heading into an AI Cold War? I'm wondering because despite the @OpenAI and @anthropic-ai CEO's calls for slowdowns, Trump and Xi Xinping have…
Would You Trust Apple AI With Your Health History? #ai #siri #apple #appleintelligence
Apple's new Siri Health features could do a ton to make your health data useful - and a lot of it is done on-device. Would you let them look at your…
The New Siri Can Summarize My Conversations. I Might Let It. #ai #apple #siri#appleintelligence
Apple's upcoming Siri Recap can summarize conversations you choose to let it hear. I'd never let any other AI do this, but I might let Siri. Why? I'll…
AI Companies Are "Gambling With Our Lives" Apparently #ai #singularity #agi #openai #anthropic
"Gambling with our lives" - that's the accusation from an Anthropic researcher that quit today. He's acting on his warning, but will that actually change…
He Used AI to Fix His Shower Drain. Then It Gets Wild. #aicoding #codex #chatgpt #gpt6 #ai
A custom shower-drain part, a winged AirPods case, and a robot prototype plan, all made with help from GPT-6 Astra. Crazy!
OpenAI want an AI slowdown... Who's going first? #aicoding #claudecode #ai #agi #singularity
Today OpenAI's chief scientist John Pachoki dropped this little gem:
Fable 5.1 vs. Deepseek V4 Pro and GLM5.3 was NOT what I expected #fable5 #deepseek #glm52 #aicoding
I was pretty shocked when I put Fable 5.1 up against Deepseek V4 Pro, GLM5.3, GPT5.6 and GPT6-Astra
I Put Fable 5.1 Up Against GPT6 Astra And Wow #aicoding #claudecode #codex #fable5 #gpt6 #gpt56
I did not expect these results when I had Fable 5.1 go up against GPT5.6-Sol and GPT6-Astra
I Put Fable 5.1 Up Against GPT6-Astra And More
Fable 5.1, GPT 5.6, DeepSeek V4 Pro, GLM 5.3, and GPT 6: I gave all five the same brief to build an agentic IDE for Ollama. Then I tried to actually use…