psychological maturity, longevity biotech, AI founder.
sf
Joined September 2016
- Tweets544
- Following705
- Followers901
- Likes1.1K
I’d 100% use it
I can finally delete hinge and stop pretending I go on hikes
and if it doesn’t work out I get my money back instead of getting really into the gym
idk why you people hate progress so much
Today @maryamyy0 and I are coming out of stealth with Sexbot.
A companion that's always there for you.
Built to fix the growing loneliness in the world.
We raised $5M to build it. Pre-orders are open now.
Link in the comments.
Readers added context they thought people might want to know
This is not a real product/service. It's made up to advertise this person's real company which creates launch videos.
Source: sexbot.services/kidding
revolut gave 2,000 engineers unlimited tokens and cut the ai bill 90% by switching to an open model.
so why is it only revolut? most companies with the same bill are still paying anthropic and openai full price.
is revolut missing something, or do american companies just buy american?
age of altron was ahead of its time
because wdym they already showed us a bigger and powerful model (astra) can edit a smaller model (qwen) lmao
b300 up 27%, a10 down 10%.
the squeeze is on the big chips only, if your product runs on a small gpu, december is actually going to be cheaper for you
this doesn't mean that AI isn't killing jobs. just that if you have more profit (coz AI), you can spend on hiring more to scale. no change for low adopters for now, but most likely they are gonna start declining.
We can finally say AI isn't killing jobs.
A new paper from me, @tryramp, and @RevelioLabs uses firm-level spend and workforce data across 21K U.S. businesses to measure AI's impact on jobs.
Firms that adopt AI heavily grow headcount 10% over two years following adoption. Low adopters see no statistically significant change.
my cofounder and I are organizing coworking with friends in SF near panhandle on Thursday 10am-6pm. text me if you wanna drop by!
holy shit gpt-6 went straight into another model's weights and 6x'd its math score
if this works beyond one math test, imagine giving it the small, cheap model you’re already running and asking it to make that better
RSI is here
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
half of sf lives in a studio where the kitchen is also the bedroom. we're cooked, literally
I think in 10 or 20 years the kitchen and living room combo layout will have disappeared from most homes
Cooking is the primary source of indoor air pollution and you're essentially releasing microscopic toxic gases into your living room and the rest of your house
On top of that 99% of kitchen vents don't do anything
That is, unless we invent some better air cleaning system for kitchens that actually work
I ran the numbers on a 100k-token input with opus 5.5
40 cents uncached and 2 cents if it’s already cached
now when I see “continue”, i’m thinking about the whole conversation going back with it. and whether i’m paying full price for the model to read it again
every linux server on earth, saved by changing kfree to kfree_rcu
internet is held together by two words
What a disaster atlasum.net/blog/cve-2026-98…
holy shit karpathy turned 43 million HN comments into a youtube watchlist
he filtered 43 million hacker news comments down to 6,454 videos. that's about 6,600 comments of people arguing per video that made the cut
this is the kind of shit i’d actually pay youtube premium for
anthropic showed an example where loading tools only when needed took the context from 150,000 tokens to 2,000
98.7% FEWER tokens
before deciding your agent needs a smarter model, maybe check what you're making it read
"claude, pls rewrite this legacy code, make no mistakes"
This is INSANE. Cisco’s president Jeetu Patel has given a 27-year-old engineer unlimited AI tokens to try to rewrite 98% of the company’s legacy code within 24 months, Fortune reports.
Padmanabhan Krishnamurthy thinks he can do it with a two-person team. He has just four weeks to prove the approach.
I hate sf halloween what do you mean my costume is god
I have access to openrouter from the future opus 19.5 and I’m a tokenmaxxer
