Sam Fold

223 posts

Sam Fold

Sam Fold

@SamFold

Engineer and Technologist at Accel. Ex Blizzard, https://t.co/4eQKILmOWH. London based but previously in Dubai, LA, Berlin, Paris, and New York.

London Bergabung Mart 2010
268 Mengikuti627 Pengikut
Sam Fold
Sam Fold@SamFold·
@mitsuhiko I benchmarked the several-thousand character code reviewer skill and it performed worse than the prompt “you are a senior engineer, analyze this code for issues and bugs”
English
1
0
30
3.6K
Sam Fold
Sam Fold@SamFold·
@IceSolst Bingo! Anyone boasting about lines of code as though it makes them some kind of genius is letting everyone know that they were never a very good engineer and their opinions can be ignored safely.
English
0
0
5
1.2K
solst/ICE of Astarte
To all the young folks learning about CS/tech, please keep in mind: Garry Tan is a fucking idiot, listening to anything he says will derail you
solst/ICE of Astarte tweet media
English
82
132
3.1K
110.1K
Sam Fold
Sam Fold@SamFold·
If anyone ever boasts to you about how many lines of code they wrote in a single day, as though that’s something special, you can happily terminate that conservation, because they’re an idiot, and neither were or ever will be, a good engineer. LOC doesn’t mean anything other than lines of code. It’s not a measure of value, intelligence, or merit.
English
0
0
0
55
Sam Fold
Sam Fold@SamFold·
@garrytan GStack is statisically worse than not using GStack. I actually proved it by running several benchmarks. The prompts make Opus dumber, not smarter.
English
0
0
19
1K
Garry Tan
Garry Tan@garrytan·
I open sourced everything about how I'm doing it. You can try it. It's free. If it works for you great. If it doesn't work for you, tell me what went wrong and I'll give you a money back guarantee. :-) github.com/garrytan/gstack
English
15
6
127
40.2K
Garry Tan
Garry Tan@garrytan·
Absolutely insane week for agentic engineering 37K LOC per day across 5 projects Still speeding up
Garry Tan tweet media
English
353
32
872
2.5M
Sam Fold
Sam Fold@SamFold·
This prompt: "You are an expert code reviewer. Review the provided diff for bugs, security issues, race conditions, and other problems a senior engineer would catch." is more effective as the 500+ line Garry Tan, "gstack" prompt...
English
0
0
0
49
Sam Fold
Sam Fold@SamFold·
@pmarca Did you not read Marcus Aurelius’ Meditations then ? Or do you not consider Marcus Aurelius to be a great man?
English
0
0
0
26
Marc Andreessen 🇺🇸
It is 100% true that great men and women of the past were not sitting around moaning about their feelings. I regret nothing.
English
2.8K
1.4K
17.6K
8M
Sam Fold
Sam Fold@SamFold·
@pmarca Except there’s no actual science in this book at all, just one guys opinion.
English
0
0
1
71
Sam Fold
Sam Fold@SamFold·
@smokeymitchell @Beefeater_Fella lol no worries :) and yeah I was wondering why my age and number of followers was relevant to my opinion on Dubai! (Especially because I was literally just there)
English
1
0
1
98
smokeymitchell
smokeymitchell@smokeymitchell·
Hey there Sam. Try not to be offended by the fact that I use the word kid. A. I’m still older than you. B. I was actually trying to respond to the Paul Trent who called you a kid. C. I didn’t look at your profile to figure out how old you were. It was my sarcastic remark towards somebody who was belittling your perspective.
English
1
0
0
10
Beefeater
Beefeater@Beefeater_Fella·
The Dubai Airport is witnessing the largest influx of departing passengers. *This post would result in imprisonment if posted from Dubai. But it’s posted from London - so fuck ‘em. 1/2
English
132
910
7.4K
1.3M
Sam Fold
Sam Fold@SamFold·
@tobi Assuming the benchmarks are any good… which is a lot harder than it seems
English
0
0
0
120
tobi lutke
tobi lutke@tobi·
Autoresearch works even better for optimizing any piece of software. make an auto folder, add program.md and bench script, make a branch and let it rip.
English
70
66
1.7K
121.8K
Sam Fold
Sam Fold@SamFold·
@garrytan Did you write this with ChatGPT, "it's not this, it's this!" ??
English
0
0
0
10
Garry Tan
Garry Tan@garrytan·
Boil the Oceans You know the phrase: “don’t boil the ocean.” Everyone’s said it in some overly ambitious meeting. It’s good advice in normal times. It keeps teams focused. It prevents scope creep. But we are no longer in normal times, and I think it’s time to retire saying it. Artificial Superintelligence means it’s time to boil the ocean. We’ll start with a few lakes first. I was recently with a university endowment’s head of private investing who told me their engineers were terrified for their jobs after seeing what Claude Code could do. And I get it — that’s the natural first reaction. But it’s the wrong one. It’s a zero-sum reaction to a positive-sum moment. Instead of worrying about doing the same thing we’ve been doing for cheaper, why not focus on doing the thing we never even dreamed of doing? Why can’t that endowment achieve 50% net IRR instead of 10%? Why can’t a startup deliver a service that is 100x better than the incumbent? Why can’t we have fusion energy? Why can’t we talk to every single user and have a perfect understanding of every bug in our product? These aren’t rhetorical questions anymore. They’re engineering problems with paths to solutions. Here is what I think is actually going on with the fear: our fear of the future is directly proportional to how small our ambitions are. If your plan is to keep doing exactly what you’re doing, then yes, a machine that can do it faster and cheaper is terrifying. But if your plan is to do something dramatically bigger, then the machine is the best news you’ve ever gotten. If you’re a worker — someone who trades labor for a living — this is the moment to become a builder. Start a business. And if you’re already management or capital, it’s time to go 10x more hardcore on what your aspirations could be. Not eking out 5% efficiency gains. Not increasing profit margins 2% by lowering cost and firing people. Those are the old games. The new question is: what would it look like to build a product or service so good that people would happily pay 10x what they pay now? The net result of this is more jobs, not fewer. As Ryan Petersen likes to say, the human desire for more things is absolutely limitless. We can actually fulfill that desire now — if we have the agency to prompt it for ourselves. Buckminster Fuller coined the term “ephemeralization” in 1938: doing more and more with less and less until eventually you can do everything with nothing. His entire vision of progress was about technology enabling radical expansion of human capability through dematerialization. He traced this from stone bridges to iron trusses to steel cables — each iteration stronger, longer, lighter, cheaper. He wasn’t describing job destruction. He was describing civilization getting better at being civilization. This is Jevons Paradox for everything. When you make a resource dramatically more efficient, you don’t use less of it — you use vastly more. Steam engines didn’t reduce coal consumption. They made coal so useful that demand exploded. The same thing is about to happen with intelligence, with labor, with every service and product we can imagine. But Jevons Paradox doesn’t activate on its own. It requires capital and management to actually raise their ambitions — to boil lakes and oceans instead of drowning them in committee That’s what startups have always been good at: moving fast in the face of radical uncertainty, building for the 10x future while everyone else is optimizing for the 1.05x present. Time to start.
English
170
296
2.2K
519.5K
Sam Fold
Sam Fold@SamFold·
@tobi What benchmarks are you actually running it against? I looked at the code for QMD... it looks really bad/amateurish.
English
0
0
0
208
tobi lutke
tobi lutke@tobi·
OK this thing is totally insane. Before going to bed I... * used try to make a new qmdresearcher directory * told my pi to read this github repo and make a version of that for the qmd query-expansion model with the goal of highest quality score and speed. Get training data from tobi/qmd github. * woke up to +19% score on a 0.8b model (higher than previous 1.6b) after 8 hours and 37 experiments. I'm not a ML researcher of course. I'm sure way more sophisticated stuff is being done by real researchers. But its mesmerizing to just read it reasoning its way through the experiments. I learned more from that than months of following ml researchers. I just asked it to also make a new reranker and its already got higher base than the previous one. Incredible.
Andrej Karpathy@karpathy

I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then: - the human iterates on the prompt (.md) - the AI agent iterates on the training code (.py) The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc. github.com/karpathy/autor… Part code, part sci-fi, and a pinch of psychosis :)

English
122
243
4.8K
794.5K
Sam Fold
Sam Fold@SamFold·
@Microinteracti1 @clashreport He didn't say that at all. You seem to be totally misrepresenting this situation. UK boots would never be on the ground, nor would we be bombing Iran, it would be using our interceptor capacity to help friendly nations with whom we are historically close, like Oman and the UAE!
English
0
0
1
207
Gandalv
Gandalv@Microinteracti1·
@clashreport The man who helped walk Britain into the Iraq hellscape is now saying the UK should do it again, only this time with Iran, a country bigger, stronger, and far more capable of turning the region into a furnace.
English
16
31
399
14.8K
Clash Report
Clash Report@clashreport·
Tony Blair on Iran: Every single time you test an alliance you never test it when things are easy. You test it when it’s hard. They were asking to use our bases to refuel. It’s not like it was in Vietnam, not like the Iraq campaign where we had thousands of British troops. The American relationship matters. It matters particularly today. It’s not a question of whether it’s this president or that president. If they are your ally and they are an indispensable cornerstone for your security, you had better show up.
Clash Report tweet media
English
674
1.8K
11.5K
791K
Inevitable West
Inevitable West@Inevitablewest·
🚨BREAKING: More than 70,000 pro-Trump Iranian patriots just CRASHED a pro-IRGC rally on the streets of London It looks like the Islamists got chased out. London stands with the TRUE Iran! 🔥
English
495
4.5K
24.2K
438.5K
Basil the Great
Basil the Great@BasilTheGreat·
🚨BREAKING: UNION JACK AND ST GEORGE'S CROSS LABELLED "TOOLS OF HATE" BY UK GOVERNMENT This is disgraceful 😡 A UK Government document focusing on "social cohesion" has BRANDED the Union Flag and the England flag as "tools of hate" RAISE THE COLOURS 🏴󠁧󠁢󠁥󠁮󠁧󠁿🇬🇧 Starmer is a traitor
Basil the Great tweet mediaBasil the Great tweet media
English
2.9K
11K
35.1K
970.8K
Sam Fold
Sam Fold@SamFold·
@owenjonesjourno @omid9 Nonsense. Nothing is proved yet. Did you actually read any of these articles you cited?
English
0
0
3
463
Omid Djalili
Omid Djalili@omid9·
Nothing is proved yet Zack. The fact that you are spreading misinformation is appalling for someone in your position. To date no one has been allowed to see any evidence of casualties or even the site itself. The regime says it’s America or Israel. The Americans say they were nowhere near the area. Here’s what Iranians inside Iran say: IRGC blew it up themselves to fill it up with dead bodies of children they’ve kept on ice hidden away in cold storage from the January slaughters to put on display in order to blame their aggressors. The fact that we have not seen the evidence yet is a red flag as we know the regime would have gleefully paraded those images on television and all over the internet. Most probably they were caught on film transferring the bodies hence why there is now a media black out on this story, as well as a full internet blackout. As Iranians we know how this regime operates. We’re talking about new levels of evil. So any narrative like the one you are pushing Zack actually supports the regime in their “don’t care how many die as long as we stay in power” quest. For the love of God and humanity, please stop. #IranMassacre #IranIsraelWar
Zack Polanski@ZackPolanski

Israel killed 160 children in a strike on a school. It's so dystopian and grim. Even worse that the BBC look to murky the waters.

English
2.3K
3.3K
16.5K
2.1M
Sam Fold
Sam Fold@SamFold·
@Fahadnaimb Very cool, but not manta rays. Manta rays look very different to this and also this is not their habitat at all, there are no manta rays in Dubai or the UAE.
English
1
0
12
11.7K
Fahad Naim
Fahad Naim@Fahadnaimb·
Dubai beachfront.... people on balconies watching Manta, others chilling near the shore, traffic busy as usual. While the news is full of drama, on the ground here it feels surprisingly normal.
English
90
139
2.6K
3M