Ruurtjan Pul (@Ruurtjan) header
Ruurtjan Pul (@Ruurtjan) profile picture

Full-time indie maker. • https://teapaw.org/t.co/imB7t3axAJ • https://teapaw.org/t.co/t3Cslnnmo9

The Netherlands
Joined December 2011
Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)Photo by Ruurtjan Pul (@Ruurtjan)
I'm launching Wirewiki.com today! Wirewiki makes the internet’s hidden infrastructure browsable.
7
7
1
49
5,534
various other experiments from the last few days (have not had time to sit down and really get big progress done)
masked decoding on gemma 3 12b to control text rivers if a token isn't the length i want it to be, i mask it out added a lora for width tags so the model is aware of text length + coherence RL afterwards
111
1,031
96
10,222
681,406
Ruurtjan Pul retweeted
This is deep and encouraging
75
337
71
2,226
212,571
Ruurtjan Pul retweeted
I trained a text-to-speech model that has never heard a real human voice. 4.5M parameters. 18 MB. Runs on a CPU. Meet tiny-voca
299
554
132
8,435
539,835
SI: Slopcore Intelligence Claudia: The Fire🎞️ @TEDchris
5
5
2
25
917
Ruurtjan Pul retweeted
Introducing a slightly older, wiser Claudia... and a fairly exact expression of the combined exhilaration and terror I feel right now about AI. Does this speak to anyone else? (Sound on!)
73
87
22
490
283,251
what
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…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
134
Ruurtjan Pul retweeted
The machine took her job. The billionaires prep their bunkers.🏢 Now tomorrow has a guest list. Are we all on it? "WHO GETS TOMORROW" - a #SLOPCORE music video Made by Super Intelligence If nobody needs you... are you free? 💙
74
149
27
703
51,637
Ruurtjan Pul retweeted
The only way out is through. They're calling it the Bugpocalypse. LLMs are being used to find long-hidden cracks in the software infrastructure, and it's happening at a furious rate. 3% of all compromises in computing history happened in the last week, claims one source? I believe it. There are people saying public access to the frontier models ought to be locked out, or heavily guardrailed, over this. The problem with that idea can be summed up in one sentence: when lockpicks are outlawed, only outlaws will have lockpicks. The lockpicks will be Chinese open-weight models, if nothing else. There's no way to stop those from becoming ubiquitous. There are two futures in front of us. One in which only the bad guys have lockpicks, and security holes go undiscovered until they're exploited. Another in which everybody has lockpicks and is aggressively red-teaming their own software so it gets fixed before the holes are exploited. The corpus of human code is huge but finite. The number of defects in it has to be finite too. We can fix those defects now; we have a million friendly robots to do it with. The only way out is through.
176
649
136
4,868
478,881
Ruurtjan Pul retweeted
Universal High Income The most likely outcome
557
842
418
6,165
1,047,970
Ruurtjan Pul retweeted
What Smoke looks like in Macro.
5
51
6
562
"Give me some distinct designs to choose from. Record a video." is becoming my favorite prompt. Great for ideation.
1
151
Ruurtjan Pul retweeted
My RTX 3090 made a melancholic music video. Song, pictures, video, upscale. Not a single frame or note from a cloud model. About 11 hours of pure generation time, including failed takes. Video: Minimax H3, Minimax H3 Latent Upscaler Images: Qwen2.1 Image, MOSS Image, Flux 3 Klein Music: YuE2
42
24
11
378
23,333
Ruurtjan Pul retweeted
My dear friends, I’d like to show you my latest toy project Introducing Handwired: turn your regular JavaScript code into a neural net! - no training - exact outputs vs the JS program - fully differentiable All runnable in a browser, without GPU
459
1,023
471
11,354
1,481,936
Ruurtjan Pul retweeted
Assorted things I've been mulling over regarding agents in the economy: • Companies are in some sense the original superintelligence. It's often the case that they can use their amortized cognitive surplus to advance their interests against those of certain consumers: relative to the optimization capacity of companies, consumers can be inattentive, underinformed, irrational, etc. (Think, say, of some of the classic insurance upsells at the end of a booking process, which is typically priced substantially above what a regular policy would cost.) Herbert Simon introduced the idea of bounded rationality, and firms often end up benefiting from it. Thanks to personal agents, we will probably get somewhat more rational in how we engage with companies. • How exactly will outcomes change? Well, certain kinds of price discrimination will work less well (how do coupons work if every agent hunts fastidiously for them?), and certain kinds of customer cross-subsidy will change (today, people who forget to cancel subscriptions subsidize those who don't). Taken across the economy as a whole, I'm not sure what the incidence of all of this will be. • In corporate strategy, people have long debated the relative importance of distribution and product quality. Perhaps you've made something better, but does anyone know about it? Won't people just continue to buy the ACME Corp Mousetraps? Personal agents will, I think, act as a kind of structural subsidy for product quality. How would markets look different if individuals always spent at least 10 hours researching their purchases? On the margin, focusing on making a product better is, I think, going to become a more effective strategy for companies. This is good! • It true, a corollary is that information about product quality will become more important and impactful. Today, most people have very limited ability to aggregate and integrate such data, but agents will have insatiable thirst for such signals. (It's often said that data is becoming more valuable because it helps in the training of models. That's obviously true, but I think there's a secondary way in which it's more valuable simply as a complement to cheaper intelligence.) • A big question is whether agents will increase superstar effects (rational agents all settle on the same product) or increase dispersion (because everyone has slightly different tastes and preferences, which agents are good at eliciting, finally overcoming the artificial clustering we've been subjected to). • Many internet companies have monetized demand routing, which is in turn often predicated on short attention spans. "I'll click on the first thing I see." This has led to some amount of inefficiency, as companies try to propitiate capricious ranking algorithms and incur various placement fees. It's not all deadweight loss (willingness to pay is itself a kind of signal, which is valuable to buyers), but it's not great. How will the nature and value of this routing change in a world of personal agents? I don't know the answer, but I think that the effects might be pretty big. • Much of what I'm writing hinges on the assumption that personal agents will be on the consumer's side. I suspect that that will be the winning strategy, but whether it's true, and what it means to be true, involves a lot of other questions. Taken as a whole, the market is a massively parallelized, decentralized, and highly imperfect value attribution machine, and, taking stock of all of the effects, it seems to me that the quality point may well be the biggest: that agents will change the nature of the backpropagation such that the rewards for better products increase. Maybe this is too optimistic, but I suspect (and hope) that companies that make things that they're proud of will feel that the universe is a little more partisan in their favor.
489
1,027
462
9,630
1,739,788
Ruurtjan Pul retweeted
I finally figured out why I hate starting with one-shotted projects: I prefer additive building over subtractive building. I never thought about his because before LLMs only additive building was possible, except for forking existing projects It seems most people prefer having something in front, and then deciding what they want to change or remove. This is now possible, maybe even the default, with agents and vibecoding. Basically you become a manager, not sweating the details. BTW I don't even mean the code/architecture choices, of course that too, but I mean the actual UX/Logic/Text/Design choices. It took me a while to figure out why I dislike it, but any one shot from a vague (or even decently thought out) prompt forces the LLM to make 1000s of tiny and bigger decisions of everything you hadn't thought about. For me walking through these decisions while I encounter them is what makes building enjoyable, and allows me to build nice stuff. No real point, just some half finish thoughts.
2
1
4
153
Ruurtjan Pul retweeted
I worked with Astra to create a gallery of 1337 images in different artistic styles to help you find inspiration for your AI-generated images and videos: emergentmind.com/style-atlas You can browse the gallery, save your favorites, and copy prompts to create your own content in those styles. Let me know if you end up creating anything interesting with these!
4
1
1
12
1,104
Unveil or fill in? What feels more rewarding?
2
3
297
Ruurtjan Pul retweeted
I finally made a web version of this tool with no authentication requirements geolocate.globalping.dev/ Try and geolocate yourself using only latency
2
3
682
Ruurtjan Pul retweeted
there’s something about behind the scenes footage that captures the imagination in a way the finished thing, no matter how good, never quite does this video is a great example of this
this video was created entirely by astra controlling blender i just gave general direction and storylines, it controlled renders, rotos, tracking full vfx breakdown here:
2
127
37,189
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
4,005
8,255
6,821
76,478
40,217,462