@theo@jaiyagill@grok Like most if not all are tiny guys who need to overcompensate. All of sudden it makes sense ... I am 7' (2m13) .... might as wel give up on my dreams and hopes, haha
GitHub Copilot costs are up 10-50x for some devs. Everyone's angry at GitHub.
Wrong target. The bill isn't the surprise. The surprise is finding out what your workflow was actually worth all along.
https://t.co/aYBdDa3rUh
#GitHubCopilot#DevTools#AITooling
Hackers took over the Obama White House Instagram this weekend by asking Meta's AI support bot to reset the password. No exploit. No breach. They didn't even need the victim's email.
https://t.co/AUrvsiqXLR
#AIAgents#AppSec
Uber burned its entire 2026 AI budget in 4 months. Their CTO: "no link yet between AI tokenmaxxing and shipping products."
Amazon built a token-usage leaderboard. Employees gamed it. Obviously.
Goodhart's Law in action.
https://t.co/JRwx5Ng9lz
#GoodhartsLaw#AIAdoption
@ThePrimeagen warned AI tools make you stop checking. His followers ran a poisoned skill from a tweet without reading it. Because he told them to.
He was right about the failure mode. He just demonstrated it himself.
https://t.co/SCO1lE3HhI
#AISecurity#MCP#SupplyChain
@theo Besides not being able to write because it did not read yet, I've been seeing 'Cancelled: parallel tool call Bash' a lot, it seems to be sending new commands faster then it can handle them ><
@theo The benchmarks aren't the real story. The real upgrade: 4x fewer unflagged flaws in its own code. A model that knows when it's guessing beats two points of benchmark every time.
https://t.co/z1SjkbJhlF
Opus 4.8 beats GPT-5.5 and Gemini on SWE-bench. That's the least interesting thing about it.
Real upgrade: 4x fewer unflagged flaws in its own code. A model that knows when it's guessing beats points of benchmark every time.
https://t.co/z1SjkbJhlF
#AICoding#ClaudeCode
An attacker wrote malware to rob Claude's workspace, then leaked his own GitHub token inside it.
Both sides vibe-code now. Your agent is the target: your keys, your installs, your configs.
https://t.co/9YgxJrqUPi
#AIAgents#AppSec
11,000 tokens spent before I typed a single word. Skills, MCP tools, hooks, all auto-loaded, none of it had done any work yet. The plugin that was supposed to save context is the first thing eating it.
https://t.co/5oeRrZtZHe
#ClaudeCode#AITooling
96% of devs don't trust AI-generated code. 48% verify it. That gap is your next incident.
The fix isn't better prompts. It's a workflow.
https://t.co/WAyUPsMsnf
#AIAgents#SoftwareCraft#AgenticEngineering
Your rate limit has a power grid inside it.
@ThePrimeagen saw the subsidy era ending. @theo saw the GPU shortage underneath. Neither went deep enough: Google's grid connection queue is 10+ years in some regions.
https://t.co/MO8nQbyF6X
#AIInfrastructure#DataCenters
@ThePrimeagen It is easy enough to ask AI to setup all the right tooling to debug though ... I must admit, there have been times i was too lazy and did just that :D
700 git tags across 4 Composer packages, rewritten in 90 minutes. v1.0.0 now points at malware. The attacker didn't publish a new version. They moved yours.
The pin didn't fail. The pin moved.
https://t.co/2qVnGZYfXw
#SupplyChainSecurity#AppSec@SocketSecurity
63k stars: plugin that makes Claude talk like a caveman.
15k stars: plugin that sandboxes tool output behind FTS5.
Output is grams. Tool output is kilos.
Pick the wall, not the manners.
https://t.co/eCuN6nLl36
#AIAgents#ClaudeCode@julius_brussee