pardal.eth (@pardal_eth) header
pardal.eth (@pardal_eth) profile picture

Interest in Ethereum, NFTs, Tesla, EVs, Renewables, Gaming

Metaverse
Joined September 2015
Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)Photo by pardal.eth (@pardal_eth)
pardal.eth retweeted
AI interfaces nowadays are still pretty much turn-based. In my PhD thesis, I thought about what interaction would look like if we could go beyond taking turns. One example is intra-turn interaction:
309
619
224
8,577
819,678
pardal.eth retweeted
Despite everything, I don't think I've ever got as many supportive messages. I won't be able to respond all, because I'm deeply focused. So I just wanted to say I'm thankful for all the kindness in the world. Also, thanks for these who invested in us against the tides. I can't promise our approach will scale, there is a graveyard of talented people who tried and failed in the past, but, if there exists a system capable of reasoning without the drawbacks of gradient descent, we'll use every resource available to find it. Meanwhile, auto-research loops keep going. HVM5 is now shockingly fast; about 4x faster than HVM4, and it runs on CPUs and GPUs via Bend's parallel lets. It is beautiful so I must resist my own temptation to open source it, as it is a key component of our SAI research. We also made a lot of progress on Bend since Studio Vibi has merged with HOC, specially on the user-facing side. More on that later...
34
18
2
905
27,780
pardal.eth retweeted
we are entering a renaissance in programming language and runtime research, and a golden age of software testing and verification it’s time to rewrite all software
536
672
189
10,231
486,601
pardal.eth retweeted
when bend2 launched, some complained about AI usage. in my view, this is only a problem if bugs can occur. the current typescript compiler, for example, has many. I hate it. with well-designed laws, though, that ceases to be a problem. so, for the better or for the worse, bend3 will be 100% AI-written. there is no going back while I was raising, I left a fleet of agents working on bend3, a small scale version of this vision. this was a huge success, beyond my own expectations. opus 5.5 successfully rewrote all of bend in itself, including its compiler, which is now certified, with zero bugs, and already 2x faster (!!). we now have a model of Bend, C, CUDA and the compiler, all written in Bend itself, and properly formalized. with this, we'll soon start the largest proof-driven compiler self-improvement loop in history news on Bend have been slow because I was busy raising, but as soon as this is over, expect updates at a breath-taking place. this is the most exciting thing I've ever done and I can't wait for the big run to start
85
63
23
1,480
49,560
pardal.eth retweeted
I finished building the interactive 3D book on GCSE and IGCSE Physics Electricity with Opus 5.5. It now has all seven lessons, from charge and current to mains and the plug. Publishing soon on my website.
I am building an interactive 3D book with Opus 5.5 on GCSE Physics Electricity. Here’s a glimpse of its lesson on mains and the plug.
101
330
29
2,830
118,693
pardal.eth retweeted
If you’re going to make something with AI, make sure it’s 10,000x better than anything you would have made yourself. If not, you just come across as lazy not smart :)
33
28
15
548
139,993
pardal.eth retweeted
i’m now fully convinced that chatting with agents is absolutely the wrong UX watching my son learn to create things, he prefers literally any other tool than prompting AI he likes to write, draw, and animate, all with his fingers on an ipad and he likes to recording real sound effects to use in his games he could have gotten much better results by chatting with an AI, but he hates the wait and doesn’t feel the result is “his work” we need a complete rethink
216
26
57
844
93,644
pardal.eth retweeted
turns out humans were putting slop inside the computer all along. the artisanal code argument never held up.
25
62
10
875
18,472
pardal.eth retweeted
Replying to @allisonology
power costs something, your attention costs something, the computers cost something
2
1
5
73
AWS AI-dlc package. Too much or the right approach? 🤔
22
let me tell you why saas is not dead and how deep in the rabbit hole we are on x ⬖ my wife is technical enough, worked as QA before, understands the concepts of cloud, hosting, frontend vs backend, etc etc. ⬖ she has 2 subs to copen ai and claude, uses them on desktop and on mobile ⬖ she has heard me rambling about ai for 4 years non stop and knows most terms ⬖ she has vibed a few websites for fun and is a creative person overall so i see she subscribed (recurring lol) to a "learn driving tests" app i asked her wtf why dont u vibe yourself one?? it will be even better, more visual, tailored to you, etc etc. she was like "uhm but why, someone already put all the effort and i just wanna be done with my driving tests" now think about the people who have NONE OF THE bulletpoints above lmfao 💀 put a price tag and just put the saas in the bag lil bro no one is vibecoding shit
209
81
68
1,934
123,800
Hmm Unikernels, Nix and their overlays 🤔
20
pardal.eth retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel. - Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all. - Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract. - Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code. - And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while. The funnel is driven by a combination of factors: 1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel. 2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel. 3. Access. Free-tier, paid-tier, internal. So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
2,271
5,951
2,303
44,420
7,060,370
Just configured dual boot Debian 13 to run Bonsai 2. Ctx 80k. 80tok/s no drafter yet. On RTX 5080
1
1
756
pardal.eth retweeted
honestly, i keep thinking about this result and it's just absurd. in the 2022 PISA, el salvador scored 343 in math, 365 in reading, and 373 in science near the bottom of PISA, close to guatemala and dominican republic now the country has developed a program for education modernization in a pilot group of public schools with 1:1 AI tutoring and personalized learning after the pilot had been running for BARELY OVER A YEAR, students got results comparable to germany and sweden. that's more than an entire standard deviation that is fucking insane we should be talking about absolutely nothing else if you actually care about poor children and equality of opportunity, the possibility of giving every child on earth an individualized tutor for a fraction of the cost of human 1:1 tutoring should be one of the most exciting developments in human history. "it's all in the environment tards" should be drooling at the news every government on earth should be running massive randomized trials of this right now, and if the effect survives, scaling it as fast as physically and humanly possible for every child on earth the funniest part is that while the supposed "fascists" are celebrating poor kids learning, the people supposedly obsessed with inequality are barely talking about it, and if anything are banning AI in schools revealed priorities are funny like that
“Results in reading, mathematics and science were above the national average and comparable to the average results in Germany and Sweden.” These are the results of our pilot program, which includes AI tutors and has been running in 171 public schools for just over a year. It was evaluated by PISA for Schools in June of this year. Since then, it has already expanded to more than 1,000 schools, and within 18 months, it will be in every public school in our country. In a few years, El Salvador will inspire the world. Security was just the beginning 🇸🇻
252
1,388
193
9,462
1,044,024
i have a degree in electrical engineering i dont mess with hardware for the same reason i dont mess with linux it’s a pain in the ass… but not any more! i can talk to my ai harness on my omarchy linux machine out loud while i work… and i gave it vision so it can see my circuits and give feedback… and it can ssh into my controller and install/run code… and it can ship prints to my 3d printer… right here it is controlling a laser and taking test readings and telling me how to adjust the sensor in realtime… it feels like magic!!!
127
89
24
1,551
59,046
pardal.eth retweeted
Replying to @DanielMiessler
Hey Daniel, this is the entire thesis of my work. You should start here openai.com/index/harness-eng… and take this to its fullest logical conclusion, which is most easy to do if you simply do not permit yourself to be in the loop while the agents work.
6
1
30
3,168
pardal.eth retweeted
Replying to @theodorvaryag
can’t find the original sorry to whoever i ripped off this meme idea from
1
1
7
73
Still trying to grasp in practicality, on how to make the Harness around the business client artifact
I distilled myself and my work to twelve theses of harness engineering. This represents the last year of my work. This is the practice and technique. Go nuts fam. github.com/lopopolo/harness-…
45