Andy Thomas

Marklet is a new sovereign #localai client for those who can maintain autonomy in the face of convenience. Under development. This is alpha v0.2.

I lived in England. We called the larger country "Britain" or "Great Britain" as in the Greater British Isles. I didn't know a great deal about NI, sorry, but why should I? My dad worked in a factory. We put GB stickers on our cars. When I was about 15 or so, a man came from the government to explain how to fill out a passport application. He said that our nationality was "British", and not English or UK or anything else. There was this thing called the "EEC". We were told on the tele, ad nauseam, that this was only a trade organisation and not any kind of union. Then one day the we just YouKay, and we were not British anymore, but just YouKay citizens and members of the EEYou. And everything about Scotland, Wales, Ireland was good, and everything about England and English people were bad. Things had just always been that way. And I like Ireland. I used to live there for while. And Scotland. My surname is Welsh. And, in the end, all the countries and all the people get ground down to be integrated into something else. It was just about order in which it done.

@LundukeJournal I was born in England, a part of Britain. One day, we woke up to find ourselves in the YouKay. I don't know where this is.

The United Kingdom has demanded that 4chan pay a £520,000 fine for failure to comply with UK age verification laws. However, since 4chan is based in the USA, the UK has no jurisdiction to fine Americans in America. “As has been explained to your agency, ad nauseam, the United Kingdom lost the American Revolutionary War. We are not in the mood to discuss the matter further, and have not been in the mood for 250 years.” The letter to the UK’s Ofcom ended by suggesting that “maybe, you could just stop sending Americans stupid letters and acknowledge the sovereignty of the United States.” 4chan’s attorney, @prestonjbyrne also included a picture of a giant hamster dressed as Godzilla.

I'm currently on Debian, Gnome with kernel 6.12. I have an AMD RX9070. This, against my initial expectation, works fine with 6.12 (although I had to upgrade the bios). I'm not a gamer, but am using the GPU for AI. Likewise, I don't have skin in the Wayland vs Xlibra debate, although I don't like how Wayland has been forced. Every now again I indulge in a bit of distro hopping. Last time it was between Fedora, Alma and OpenMandriva, but ended up back at Debian. In the future, I may look to a BSD variant.

@kuiperzone According to our users, the current state of AMD in XLibre is good. Which distribution would you like to switch to?

The YouTube channel "I use Devuan btw" has been testing #XLibre with the #Nvidia RTX 4060 using CUDA driver 590.48.01. He says, "It works great. OS is #devuan 6 excalibur with #openbox window manger. If your gpu supports version 570+ it'll work." #XingsDay youtube.com/watch?v=gbhtNW…

I agree. I have been articulating that AI will destroy the willingness to create since 2022, but with little traction, although things are changing now. I've come accept that nature is going to take its course. I don't think the plethora of AI generated content will persist. I think those who fall into it will atrophy. In the end, there will be a few who will be a able to maintain autonomy in the face of convenience and for them, just perhaps, AI may serve some limited purpose. I could be wrong. For my own project, I just decided that I would do it anyway and keep doing it, even I know/suspect it will be "grifted" by AI engines. It is hard to keep going in the face that. This goes outside physics, but it actually helps me to think that it isn't "mine" anyway, but a privilege to do it and to let the universe decide whether it should persist or not.

@kuiperzone @JoshuaLisec But it's important to be mindful of how our society functions. IP is a core part of it. If anything somebody produces can just be taken, people will stop producing. The is was so important, the first US patent laws were created in 1790. Copyright and patent help enable progress.

This is an important point, but this is not legal advice: The U.S. Copyright Office guidance on A.I. use states something along the lines of . . . Your A.I. content is copyrightable if you “creatively transform” and “incorporate” the output. In other words . . .

@PyresMagicka If I understand correctly, I broadly agree. Although I am creating an AI project, the effects of AI are poorly understood. Those who use it to create a plethora of meaningless content will be washed away. Only a few will persist.

This is a discussion worth having. I used to be a freelance graphic artist. I changed career paths because my American pay rate was outsourced by thumbtack and fiverr to people in India and Indonesia who can produce equal quality for 1/10 of the price. AI generated art, Canva and similar have outsourced their work with prompts and templates. No it doesn’t allow for pixel perfect design. But 9/10 times that doesn’t actually matter unless you’re a high end magazine or marketing firm; in those cases you can afford a design team. It used to require a college degree or equivalent to use Adobe or Corel. It used to be most people couldn’t write much of anything. Now it’s different. People with visceral hatred for AI or templated art generally are seething at how the creative tools at the world’s disposal have made their own toil seem trivial because someone else pooped out 200 books and made more money last year. Nobody remembers those books. None of them will matter in 3 years. Are you writing for the ages? Do you desire to be the next to capture the soul of your nation and write its epic? I do. I may not, but I write as if I will. Use the tools at your disposal how you see fit. Focus on your art. We can do better.

@forallcurious This is an over-simplification. My understanding is that the "past" can be only altered before quantum states interact. Once they do, it becomes like hash and attempts to alter it then are simply "lies".

🚨: Quantum Eraser experiment shows that past events can be altered, leaving physics forever shaken and challenging our understanding

Hi Paul, I understand your point and it's how I would have seen things in the past. Moreover, I think if you're using Grok to "create" things for you, you haven't really created anything. My perspective is much deeper. We are all, ourselves, creations of the universe itself. Moreover, I do very much believe in free-will, but not as a "this or that" thing. I.e. you just have it or you just don't. This to me is like asking whether light is just a wave or just a particle when in fact it is something else. I subscribe to the views of cosmologist, David Layzer, in which "information" is negative entropy. I see this as very relevant to LLMs btw in that they truly create nothing new, but we can. I see my view as ultimately being the opposite of the negativity contained in your retort. However, it's a hard slog to get anyone to see this as human beings like to collapse things into a "this" or "that".

@kuiperzone @JoshuaLisec That sounds like a good way to always be poor and never inspire people to create.

@paul_e_jones @JoshuaLisec And that's a serious answer by the way.

@JoshuaLisec If Grok generates it, whose AI content is it? Serious question.

@notakeynesian @Grummz Good. Just because software is open source doesn't mean those who have written it are forced to ingest random PRs from the internet. Sorry.

Perhaps, but there's a simpler and more direct reason AI is "killing" open source. It will quickly overwhelm the microeconomic business model of the git repo providers. The average per Senior Engineer used to be 2.5 PRs/week. Maybe as many as 4.0/week if you take the high-side. Now we're quickly accelerating to closer to 10/week, and it will continue to not only grow, but at an accelerating rate. When that hits 20, 30, 50 PRs/week, then the bottleneck isn't about PR-reviews. After all, we can just have AI's doing PR reviews (ignoring the complexity of that in and of itself). The problem will be basic infrastructure economics. Github, Gitlab, Bitbucket, et. al. simply cannot earn a marginal revenue higher than their marginal costs if they're handling anywhere near those loads. They're already at thin margins now. This will kill them. Or they'll raise prices so high [more likely] that it will create a strong incentive for more fragmentation and self-hosting on more localized networks in order to avoid the teeth of the economics associated with truly "public" open source projects. In other words, we'll be back to the late-80s thru late-90s era of open source hosting. It will be mostly universities, corporations and a few private providers who themselves will "guard" participation, thus rendering it not-so-open-open-source.

I think AI has killed Open Source. In the short term, AI is going to overwhelm Open Source reviewers who won't be able to keep up with reviewing submissions. In the long term (12 months), there is no need for Open Source software when AI can write it all on demand.

@pvergadia We knew all this years ago. And no, I'm not fixing AI's damn code.

🤯BREAKING: Alibaba just proved that AI Coding isn't taking your job, it's just writing the legacy code that will keep you employed fixing it for the next decade. 🤣 Passing a coding test once is easy. Maintaining that code for 8 months without it exploding? Apparently, it’s nearly impossible for AI. Alibaba tested 18 AI agents on 100 real codebases over 233-day cycles. They didn't just look for "quick fixes"—they looked for long-term survival. The results were a bloodbath: 75% of models broke previously working code during maintenance. Only Claude Opus 4.5/4.6 maintained a >50% zero-regression rate. Every other model accumulated technical debt that compounded until the codebase collapsed. We’ve been using "snapshot" benchmarks like HumanEval that only ask "Does it work right now?" The new SWE-CI benchmark asks: "Does it still work after 8 months of evolution?" Most AI agents are "Quick-Fix Artists." They write brittle code that passes tests today but becomes a maintenance nightmare tomorrow. They aren't building software; they're building a house of cards. The narrative just got honest: Most models can write code. Almost none can maintain it.

Have read his book. I think he is on to something regarding information-mass equivalence, but suspect he has got some things wrong. I don't buy the "simulation" or "just an illusion" angle. I don't think information itself is digital. If it was, we'd live in a deterministic universe, but we don't.

PORN= "Did this graduate student just create real life teleportation?" "Scientist finds hints of infinite cheap energy thrown away as trash." "CERN physicist sees fleeting glipse of possible 'new form of matter'." "Have 3 Kansas Scientists solved Einstein's deepest mystery?"

I have been articulating this since 2022, but the response from many has been a tad depressing. This article, however, is to be commended for its clarity. linkedin.com/pulse/brave-ne… "I think the Choice Engine’s greatest danger is that it works." - Yes. This is the trap. "AI can increase agency while eroding autonomy." - An important distinction and very astute! And... "The structural effect of Choice Engines may be to concentrate the capacity for self-rule in those who already have it." I have reluctantly come to the same conclusion. I have would not have wished for this on people, but in the end I will accept their choice. Autonomy is a thing to be taken, not imparted.

@elonmusk @_kaitodev @garrytan @karpathy If you're believing this, you're already disposable.

@_kaitodev @garrytan @karpathy All jobs will be optional. There will be universal high income.

5 minutes ago, @karpathy just dropped karpathy/jobs! he scraped every job in the US economy (342 occupations from BLS), scored each one's AI exposure 0-10 using an LLM, and visualized it as a treemap. if your whole job happens on a screen you're cooked. average score across all jobs is 5.3/10. software devs: 8-9. roofers: 0-1. medical transcriptionists: 10/10 💀 karpathy.ai/jobs

@LundukeJournal Thin end of the edge. Once in place, there will be requirements for all software to use the API. And once that's in place, it won't just be about age verification anymore but all manner of things. But then again, it could all lead to unintended consequences.

Is Meta Behind the Age Verification Laws? It's true, Meta lobbied for specific age verification laws. But that is only a small part of the story, which involves Google, Apple, OpenAl... and even Roblox. The Meta Lobbying document: github.com/upper-up/meta-…

@ghostbsdproject I am happy, humble and proud to be celebrating... my 15th follower! I'm hot the heels of Musk himself here.

We have just passed 4,002 followers on X.🎉 Thanks, everyone, for the support!

Am I correct in viewing the multiverse idea as a way to explain whatever it is you want to explain in the way you want to explain it by simply proclaiming an alternate reality? Is not science the study of Reality? How do you test these alternate realities? Are invented computer models, modelling whatever it is you want things to be, considered to be the test? We have a sample size of just one, and the fact remains that in this sample there is intelligent life when, according to a mechanistic interpretation of the second law, we should not.

Theoretical physicist Sean Carroll was recently interviewed by podcaster Alex O’Connor and asked to defend his stance that one of the most thought-provoking scientific arguments for God’s existence, the argument from cosmological fine-tuning, “is the best argument for God, but it’s still a terrible argument.” Philosopher Bruce Gordon responds. 👇

@Venkydotdev The universe runs on gratitude.

STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI STOP SAYING THANK YOU TO AI