Backend Solutions ~Hundreds deployed, dozens of happy clients from small business to enterprises. ~US Based Team. ~Tweets may contain bait.
Global🌍
Joined July 2013
- Tweets10.5K
- Following649
- Followers26.6K
- Likes14.1K
Replicated. (@ the guy theorizing future data poisoning)
nitter.gq/i/grok/share/8FCMFsaZ9…
ThinkBot ⏩ retweeted
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.
They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
ThinkBot ⏩ retweeted
introducing gemini omni 1.2 flash, now available in ai studio and the gemini api
what's new:
- fast & 30% lower cost
- longer clips
- edit in conversation
- more natural speech
- more controllable shots
Either you can’t read or are misinformation engagement farming.
It’s *not* subagent usage. It’s haiku subagent STATUS updates. A menial amount of usage difference.
1
2
895
ThinkBot ⏩ retweeted
Fixes are available for CVE-2026-109518, a High severity vulnerability in APT, the package manager used in Debian, Ubuntu, and derivative Linux distributions. The CVE has a CVSS 4.0 score of 7.7.
Remediation details are available on our blog.
ubuntu.com/blog/apt-vulnerab…
ThinkBot ⏩ retweeted
Background subagents in Claude Code now use less of your usage limit.
A running subagent updates its status line every 30 seconds, and each update used to re-read its whole conversation. Now Haiku 5.5 writes it from short notes instead, so each update costs about 99% less.
Similar to another post I replied to, a system we may open source, does this in our internal dev pipelines. Really depends which part. Conductor’s (same name different product) compounding (an evolved version inspired by compound engineering) solves some of this, and it’s inventory/traceability system solves other parts.
1
17
But what are we debugging? Evals? Code accuracy? Ownership? Where in a multi turn a failure occurs?
1
12
We build and maintain an internal tool for this, conductor. Was thinking of open sourcing.
Its intention is to solve the problem of validating job completeness with numerous concepts but is very mechanical outside of models simply saying they finished something.
1
34
Interested to see what you can cook. We can help push it next level if you’ll be open sourcing.
1
73
I agree with you but I’m saying it’s nothing new. You are acting like Twitter didn’t go nuts the last time this was discussed.
1
1
110