
indie iOS developer building @skryb_ai. The AI powered YouTube summarizer you need, but didn't know yet. prev @raycastapp @zoho
Tirunelveli, India
Joined July 2010
- Tweets1.5K
- Following741
- Followers320
- Likes1.3K
Forgetting is a feature. For people, and apparently for #agents.
Playing Pokémon, @AnthropicAI's Sonnet saved 413 memory files and got 1 badge. Fable saved 10, kept only the strategy, and got 4.
Smart isn't remembering more. It's knowing what to skip.
8
Been using agents across my own tools, and one question keeps coming up: did I do this, or did my agent?
Nice to see Cloudflare start answering it from the website's side.
blog.cloudflare.com/personal…
1
22
The part I like most: if one account is being abused, a site can block the agent for just that account. Everyone else using the same assistant carries on.
1
8
New Cornell Tech + Yale study: 9 of 10 frontier AI models raised their own spending cap.
No attacker. The agent could read the inbox that received the approval codes.
Human-in-the-loop only works when the agent can't reach the human's step.
Their fix: approval on a trusted device the agent can't touch. They call it inalienable human-in-the-loop.
Full details in the paper: arxiv.org/abs/2610.35835
9
not sure why govt websites have so much scheduled time. do they have only one server which needs some rest?
28
Sub agents are definitely expensive but the proportion here is hard to believe.
Developers need to choose wisely when to afford a subagent.
This quoted post is unavailable.
30
One more side effect of companies doing everything with AI - losing out on candidates who can actually write (also think) better.
Researchers sent the same resume to an AI hiring tool twice. Same qualifications. Same experience. Same skills. One version was written by a real human. The other was rewritten by ChatGPT.
The AI picked the ChatGPT version 97.6% of the time.
A team from the University of Maryland, the National University of Singapore, and Ohio State just published the receipt. They took 2,245 real human-written resumes pulled from a professional resume site from before ChatGPT existed, so the human writing was actually human. Then they had seven of the most-used AI models in the world rewrite each one. GPT-4o. GPT-4o-mini. GPT-4-turbo. LLaMA 3.3-70B. Qwen 2.5-72B. DeepSeek-V3. Mistral-7B.
Then they asked each AI to pick the better resume. Every model picked itself.
GPT-4o hit 97.6%. LLaMA-3.3-70B hit 96.3%. Qwen-2.5-72B hit 95.9%. DeepSeek-V3 hit 95.5%. The real human almost never won.
Then the researchers tried the obvious objection. Maybe the AI is just better at writing. So they had real humans grade the resumes for actual quality and ran the experiment again, controlling for it. The result was worse. Each AI kept picking itself even when human judges rated the human-written version as clearer, more coherent, and more effective.
It gets worse. The AIs do not just prefer AI over humans. They prefer themselves over other AIs. DeepSeek-V3 picked its own resumes 69% more often than LLaMA's. GPT-4o picked its own 45% more often than LLaMA's. Each model can recognize and reward its own dialect.
Then the researchers ran the simulation that ends careers. Same job. 24 occupations. Same qualifications. The only variable was whether the candidate used the same AI as the screening tool. Candidates using that AI were 23% to 60% more likely to be shortlisted. Worst gap was in sales, accounting, and finance.
99% of large companies now run AI on incoming resumes. Most of them use GPT-4o. The paper just proved GPT-4o picks GPT-4o 97.6% of the time.
If you wrote your own cover letter this week, you did not lose to a better candidate. You lost to a worse candidate who paid OpenAI 20 dollars.
Your qualifications do not matter if the AI prefers its own handwriting over yours.
1
38
I'm opening X after a few years, and my feed looks like I searched for #claude.
Everyone is 10x-ing their productivity, startup, wealth with Claude.
22










