i build ai apps. mildly afraid of live demos.
SF Bay Area
Joined October 2026
- Tweets975
- Following139
- Followers165
- Likes330
Tutor just raised a $75M Series B and rents robots at about $16 an hour.
The part I didn't expect: the CEO says the ongoing AI inference and remote human help now cost more than the robot itself.
We raised a $75M Series B to deploy intelligent robots now, led by @a16z, with participation from @usv and @neo.
@tutorintel robots work real jobs across the United States, directed by human “tutors” and the foundation models we’ve trained on the work done. This round accelerates frontier model training, robot manufacturing, and deployment into industry.
We’re in almost 4x as many sites as a year ago; nearly a third of this year’s deployments went to existing customers coming back for more; and we’re on track to have more than 1,000 robots under contract by the end of 2027.
Our mission is to proliferate robotics for the common good. If you want to make robots reliably do things others think are impossible, for people who’ve never had access to this technology, join us.
Today's Stockfish on a 1995 PC could've beaten Kasparov, Nicholas Carlini says.
About 3300 Elo on a 200 MHz Pentium Pro with 64MB of RAM.
Most of the gain (+720 of +1335 Elo) is software. I'd love to see the same split for LLM inference.
nicholas.carlini.com/writing…
$119B went through shopify checkouts last quarter.
free cash flow margin 22%, up from 18%. revenue +32%.
stock fell 15%+ anyway because GMV growth cooled from 32% to 29%. tough crowd
shopify.com/investors/press-…
openai is raising bug bounties - new rates apply to pending reports and reports from the last 30 days.
models find bugs faster now.
honestly, paying the friendly finder more than the black market is just cheaper
As AI becomes more capable at finding vulnerabilities, defenders need to put it to work before attackers do.
We’re planning our next internal security sprint at OpenAI and partnering with Cathedral Systems, whose researchers will use our models alongside our teams to challenge our defenses. They’ll bring differentiated expertise and approaches to our ongoing work to find and fix weaknesses with AI.
We’re also raising bug bounty rewards, including more than quadrupling rewards for critical findings. Exceptional findings will be eligible for up to $172,000, including bonuses.
We’ll keep advancing our models, using them to strengthen our own systems, and expanding access for trusted defenders.
openai.com/index/preparing-o…
giving every PM an on-demand data scientist is great
the next thing i'd want: the query behind each answer, one click away. most "the dashboard is wrong" fights end up being a filter someone forgot
@poteto
if a small reasoning model looks bad at math, check how often it just loops
in the mit weight-space post, qwen3.5-0.8B hit the token limit on 93% of math answers. astra removed one loop direction from 0.016% of its weights and MATH-500 went 7.9% to 48.1%
rikiyat.github.io/blog/weigh…
if you run agents on claude, this one's in a release note, not a launch post
v2.4.0 of the python sdk adds tool_use_limit_reached_error and compaction_limit_reached_error
two annoying ways a long agent run stops. now each one has its own error type
github.com/anthropics/anthro…
openai built a life admin benchmark. kindergarten enrollment scored 93.1%, a diabetes meal plan 90.4%
the 100% gpt-6 astra runs are labeled "best recorded". i want to see the median run before it books anything with my card
openai.com/index/lifeval
sam's quote is at about the 29 min mark of his cheeky pint interview, writeup here: tomshardware.com/tech-indust…
