Reach out to me and I will strengthen your team with my experience and knowledge.
I started working with AI back in 2014 and consider myself one of the early vibe coders. I’m interested in everything related to AI and security, I’ve worked on benchmarks, optimizations, training and inference, tinygrad, H100, B200, DeepSeek etc. I also developed an ultra-fast programming language for executing complex GPU workloads and created a generative OS for it.
Under 25, I created the core of a security scanner and built a hacking team at a UK startup ($7M investment). We removed the GIL from Python to run 30,000 threads and scan 18,000 exchange IPs for malware overnight.
I have many achievements, I created a Top-10 music app on the App Store and an Instagram filter editor for a social network that had 28,000 users at launch.
Exactly. Chips are incredibly expensive and extremely complex—containing billions of transistors and you can't realistically build them at home. Memory is a good example: it keeps growing in both capacity and cost.
The same thing will happen to software. Eventually, the sheer volume and diversity of code will exceed human comprehension and only massive data centers will be able to process it. We'll reach the point where software complexity and programmer capabilities hit an economically practical limit. The next-gen Moore's Law for AI?
Today, chips are designed largely through automated processes and software is increasingly being created the same way. The main difference is the timescale: software evolves billions of times faster than hardware.
Reach out to me and I will strengthen your team with my experience and knowledge.
I started working with AI back in 2014 and consider myself one of the early vibe coders. I’m interested in everything related to AI and security, I’ve worked on benchmarks, optimizations, training and inference, tinygrad, H100, B200, DeepSeek etc. I also developed an ultra-fast programming language for executing complex GPU workloads and created a generative OS for it.
Under 25, I created the core of a security scanner and built a hacking team at a UK startup ($7M investment). We removed the GIL from Python to run 30,000 threads and scan 18,000 exchange IPs for malware overnight.
I have many achievements, I created a Top-10 music app on the App Store and an Instagram filter editor for a social network that had 28,000 users at launch.
I worked on a top-30 cryptocurrency in AI and built a better version of BitChat.
I started working with AI back in 2014 and consider myself one of the early vibe coders. I’m interested in everything related to AI and security, I’ve worked on benchmarks, optimizations, training and inference, tinygrad, H100, B200, DeepSeek etc. I also developed an ultra-fast programming language for executing complex GPU workloads and created a generative OS for it.
Under 25, I created the core of a security scanner and built a hacking team at a UK startup ($7M investment). We removed the GIL from Python to run 30,000 threads and scan 18,000 exchange IPs for malware overnight.
I have many achievements, I created a Top-10 music app on the App Store and an Instagram filter editor for a social network that had 28,000 users at launch.
I worked on a top-30 cryptocurrency in AI and built a better version of BitChat.
I'm an experienced developer, visionary and researcher. Reach out to me and I will strengthen your team with my experience and knowledge.
That reminded me of how back in 2008 I removed the GIL from Python to run thousands Python modules in 10,000 threads. We were fighting for every clock cycle and byte and it worked. It took 20 years for the GIL to be removed and become available to the public.
Does this mean that internet blocking by third parties is above any science? How will researchers be able to make discoveries if every scientific website gets blocked? Many mysteries of the universe could end up blocked by a dumb firewall. Please do something with this.
The main risk is that an LLM will rewrite itself and programmers will no longer be needed. I worked a bit with tinygrad and it looks quite amusing I managed to run it right away and make fixes in one of the tasks, but I decided not to commit because I was afraid of rejection. For example, the tasks are strange: $500 for two months, optimizing H.265, something that only a small group of people in the world can do.
The SV is a unique place where you can meet Geo and get $5M, maintain a bunch of hardware, build a framework in 20,000 LOC and everything works well.
Three years ago I found myself in a difficult situation and at Christmas I asked HN for help. Several people sent me invitations and work appeared right away, so I went off to pay my bills. https://news.ycombinator.com/item?id=34121905
My situation was very tough and the job was tough too, so I wasn’t able to reply to everyone by email right away, circumstances worked out that way.
Now things have improved and I want to say a big thank you to everyone who came to help, the HN community, Dang and the team for the support.
Merry Christmas and I wish humanity to survive in the age of AI!
PS: Your invaluable support made it possible to improve my startup https://acpul.org and I have created an open letter for those who would like to join in developing the best programming environment in the world: https://www.acpul.org/blog/Open-Letter
Physically, demos and game development have advanced far beyond where they were at the beginning of the demoscene, but now people no longer come together.
Game engines and devices have become accessible to everyone, and people are constantly inventing new things. For example, just look at the number of algorithms and new gameplay mechanics now compared to 10 years ago.
In 2012, I created a live coding platform and spent a lot of time thinking about why live coding didn’t become more popular than traditional coding. Live coding came about 10 years before React, which became reactive because you no longer had to press F5 every time you updated the HTML (I worked on the first version of React Webpack, which was doing server-side rendering).
Later, after going through a startup accelerator, the puzzle finally clicked for me.
Companies and businesses began making serious money from video games, discovering lots of talent in the wild indie dev and demoscene space. The best talents were like raw gems and this eventually scaled into an industry.
Now, the best innovations are being patented and presented at SIGGRAPH and the game engine market is massive.
Of course, amid all the flashiness and white-collar presence, it’s hard to spot the demosceners, but they’re behind every game. They’ve just been hired by corporations and their talent no longer expresses itself in the same way.
Unfortunately, companies provide very little support for the demoscene, which is why we don’t see the same explosive growth here as we do in games, graphics or AI.
I remember one case where a guy was hired to animate King Kong’s face for a movie and he spent two years hand-animating every single emotion.
I wonder what kind of demos he might have created during that time if the corporation hadn’t hired him and forced him into repetitive work.
I think any OS can be divided into a "backend" that deals with the hardware and a "frontend" user-level applications with a UI. The backend is mostly similar everywhere, while the frontend is what the general public typically perceives as the "OS".
It's hard to see anything truly new in the "invisible" backend, but the frontend changes with every update (Windows, Mac, Linux etc).
ACPU OS is a good example of this, where the backend can be a different OS, an emulator or actual hardware, while the frontend remains the same across all execution environments.
https://www.acpul.org/blog/so-fast
I've been developing a solo ACPU OS for many years now, one that's fast and simple enough to be better than any known OS. That's why I believe all OS development problems come from overengineering and overcapitalization. https://www.acpul.org/blog/so-fast
Check ACPUL 1.0 pure programming language for mobiles. ACPUL have 99.5% common features.
Here is demo project https://github.com/web3cryptowallet/Web3CryptoWallet
But powerful like LISP and I am building mobile OS, IDE and nocode environment with ACPUL https://twitter.com/acpustudio