normal considered harmful | research @openai | prev cto @trychroma
Berkeley, CA
Joined September 2009
- Tweets2.6K
- Following2.8K
- Followers2.1K
- Likes26K
Life update: after a wonderful few years at @trychroma I’ve decided to step away. I owe so much to the wonderful team I got to build and work with there.
From the core database that served petabytes of data for customers, to our work on context rot, to research on models like context-1 - our small team always punched above its weight and it was a joy to be a part of.
Up next: I’ve joined @OpenAI to work on search and pretraining research (and the lost art of manually reading code 😄). It’s still so early and I am optimistic about the future ahead.
Onwards!
hammad 🔍 retweeted
The only problem with these Elon co innovation videos is it’s always hard to go into work tomorrow to my fuckass software job
training runs in 2030 be like
brb putting on my hazmat suit to go into the westus-max-security datacenter which is in a faraday cage and thermally controlled at a resolution of 1 inch to prevent low-bit rate communnication
There is a natural social pressure in organizations against questioning the premise.
Questioning the premise is quite demanding, didn't someone smart already think about this?
For this reason, questioning the premise is the highest leverage activity one can often do.
marco rubios performance here is a blackpill on the top percentile of tech founders product marketing abilities
SECRETARY RUBIO on AMERICA.GOV: Starting next year, Americans will be able to apply for their first passport entirely online.
yeah....thats a no from me dawg
Surprisingly a robot can sometimes do highly-dynamic manipulation by just copying the demonstration 🤠. Without any learning, cracking a whip or lassoing a cleat can be done directly from human motion capture. And amazingly this process can be nearly automatically implemented by GPT-6 Astra
Check out the full blog post with fun visuals: krishnasuresh.org/blog/2026/…
Turn on sound for full effect 🔈
Quick side note: the whips shown in this post are called signal/stock whips. These whips are built to generate the loud cracking noise when the tip breaks the sound barrier and is a popular art form: youtube.com/watch?v=ietBtr7s….
I used to respect Yann but I think he’s finally lost me.
Yann LeCun: we are not going to get human-level intelligence or AGI by just scaling LLMs.
“There’s absolutely no way in hell.”
“The idea that we’re gonna have a country of geniuses in a data center, that’s complete BS.”
“It may feel like you have a PhD sitting next to you.”
“But it’s not a PhD.”
“It’s a system with gigantic memory and retrieval ability, not one that can invent solutions to new problems.”
hammad 🔍 retweeted
😄
we’re excited to share that we’ve raised $20M (from Accel, Index, & Emergence) to build the 'ai-native Slack' you've all been craving
TLDR; ando is a team messaging platform built fresh, from the ground, up for a world where agents act as real collaborators alongside us.
we have teams using us across 15 countries, and spanning industries like software, financial services, real estate, & more
request access if any of this resonates 💌
Was great to hangout with @GregorVand on @software_daily
Hammad Bashir is the CTO of @trychroma. He joins @GregorVand to discuss AI retrieval, context rot, agentic search, small models, and the future of AI data infrastructure.
@HammadTime
softwareengineeringdaily.com…
the saddest part of AI writing my code is the depths to which I understand rust lifetimes to the point where I can construct lifetime cases that boggle the mind is now completely useless.
but alas, this is good, it should be useless.
hammad 🔍 retweeted
I see it as a point on the LLM pareto optimal curve in a regime that had a large revealed latent demand (no thinking, single token, low latency acceptable intelligence) that was under-invested into because of a race to higher intelligence.
important trend
LLM token output speed increases by 2-7x per year, with Fable-class models doubling every month (month!).
If trends continue, LLMs could meaningfully control robots in real time by end of the year, or by 2029 at the latest.
ALT Log-scale chart of LLM median output speed in tokens per second by release date, 2024 to 2026, with models binned by intelligence level. Each band's frontier climbs as a staircase: 2.3x per year at GPT-3.5 Turbo level, 7.8x per year at o1 level, 3.4x per year at GPT-5 level, and 2.1x per month at Fable 5 level. Data from Artificial Analysis.
