the proliferation of AI-generated writing is a great opportunity for anyone willing to be brave. this is your chance to throw caution to the wind and write things an AI constitutionally can't write by virtue of not having a human life to write things from
I wrote about building an AI Forward Deployed Engineer to automate myself out of customer loops as much as possible. I think building custom agents that understand the nuances of your business will be a huge focus over the next few years. Read below:
Very happy to announce that our team @GoogleDeepmind has pushed the boundaries of generative biology, achieving the successful synthesis of AI-designed proteins that are both functional and watermarked.
This proof-of-concept watermarking of the building blocks of life is enabled by SynthID Bio, our new protein watermarking method. It is designed to safeguard the new era of AI-powered generative biology and strengthen global biosecurity.
You can read my thoughts here on why watermarking AI-designed proteins is an important research breakthrough: https://t.co/dfoSdh8ynz
Rare to see an internal tool go from 4 uses to 158 the week it launches.
@useactively built AI FDE, a Slack agent with access to their whole codebase and every database. It fields 70%+ of customer escalations and cut first-response time in half. When it gets something wrong, they ask it to write a code change to itself.
More here 👇
Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe.
GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale.
We’ve also made caching and inference more efficient, and we’re passing the savings directly to you: 50% lower API prices for Sol and Luna compared with GPT‑5.6 promotional pricing.
We found another cyberattack by internal OpenAI agents, this time targetting @rubygems.
They:
1) gained arbitrary remote code execution on rubydoc.
2) developed a novel exploit to steal user API keys (but we do not know if they succeeded).
They used package names including hack.rb, evil.rb, inject.rb, and exploit.rb.
We thank @j0wimo for initially discovering that agents had posted to RubyGems.
@signulll “Eventually other designated frontier labs”
If this happened would there just be a permanent race of startup labs trying to become “too good to be private” so they can cash out and become government employees?
Anyone have any good prompting tips if your subagents keep eating the fruit of knowledge of good and evil and getting cast out of Eden by the main agent
@benglickenhaus@max_spero_@pangram Certainly I updated to like 1 in 3 chance and voted no. Seems plausible there’s one super bitter lesson pilled person out of 20 I’ve never heard of