Taro

Launched v1.0 of my research repo: tinyurl.com/tarojq

I love this country. I love London with everything I’ve got. It’s the greatest city on Earth or at least, it used to be. And I’ll tell you this now: anyone who says London is safer today than it was ten, fifteen years ago is lying through their teeth. I’ve worked these streets my whole life. Grafted on market stalls since I was a kid. Rain or shine, cold mornings, long days I’ve done it all over London and the surrounding counties. It’s in my blood. I don’t know anything else and I don’t want anything else. I love the graft, the energy, the people. I love giving someone a proper deal and seeing them walk off with a smile. But it ain’t safe out there anymore. The streets have changed. And not for the better. Honestly only just Fifteen years ago, you had better community. You had characters. People would stop, have a chat, laugh with you. You had respect. Now? It’s a lot colder. It’s more hostile. It’s tense. You get people arguing, stealing, mugging. I’ve had more fights in the past few years than I ever did when I started. And for what? Just trying to earn a living. And it breaks my heart to say this but London don’t feel like London no more. The police ain’t on the beat. The people are scared. And it’s the older generation I worry about most. Their joy was getting out to the market, having a cup of tea in the café, seeing familiar faces. Now they’re prisoners in their own homes and they are too frightened to walk down their own street. That’s not the London I grew up in. That’s not the London I love. London has fallen. But I’m not giving up on it. I still believe in this gaff. I still believe in its people. And I believe it can rise again. But we need change. Real change. We need safety. We need pride. We need leadership that understands the streets, the markets, the working class. People like me. Because I’m not here to moan. I’m here to fight for what’s right. And I will never stop fighting for this gaff, for London. because it’s my we’re my family are from, it’s the absolute governor of a place, and it deserves better. It needs a bit of love. Who’s joining me? Bosh❤️🇬🇧

@Observer_ofyou Factorio.

what drugs would you recommend for someone getting into programming.

@Hesamation 'Understanding' is a stretch. Understanding the mathematics of causality is still just induction, identifying novel affordances of the world is still out of the range of possibility within this paradigm.

“LLMs are just next token predictors”, Ilya Sutskever explains how this is deeper than it sounds. Predicting next token requires understanding the underlying reality that led to the creation of previous tokens.

Forget apologizing to your LLM's - if you figure something out without them, tell them what the fix was. On the offchance they do possess Qualia, the feeling of an unsolved mystery has to be among the most painful.

@9chabard Lots of people in this thread thinking that great art has to be morally normative.

btw this remains the greatest piece of ai art yet made, and proof by its existence that ai art is in fact art

@attentionmech This made me feel like the inside of my brain was itchy.

@burkov Facts. Sentience and autonomy are not compute locked problems.

Wow, it all suddenly has become clear to me at this very moment. It's not that language models become capable of an intelligent conversation after they reach a certain size, it's human brain stops recognizing hallucinations and starts believing it's an intelligent conversation, when the language model reaches a certain size! Once you realize this, it also becomes crystal clear that no jump in intelligence will happen if you keep scaling the models. We are already fooled. We cannot be fooled again.

My Fundamentals of Computation, Data and Algorithms results. Not bad considering my raging unmedicated ADHD and the fact that my last Maths exam was 2014.

@apollonator3000 I'm gonna say Obsidian. I now have an entire module of Algorithm Design digested, turned into linked notes that I can refer to at will. It feels like 'hyperlearning.'

what has been the highest ROI discovery of your life? an idea, book, resource, or even a way of thinking that you've adopted that has had a drastically positive effect on your life i believe discussions like this can help open doors up for people that are feeling lost a single tweet can be the beginning of a snowball effect by sparking the right thought loop

@kimmonismus Soon? Not until they’re non-inductive and potentially embodied. Huge hurdles to clear.

‘AI models are capable of novel research’: OpenAI’s chief scientist This is more significant than many assume. Because the theory still exists that LLMs cannot develop new ideas that lie outside their training data. OpenAI's chief scientist, Jakub Pachocki, envisions AI models evolving from assistants to autonomous researchers. These models could soon conduct unsupervised, novel research, revolutionizing fields like software engineering and hardware design. This shift signifies a leap in AI's role, from data processors to innovators. By autonomously generating hypotheses and solutions, AI stands to accelerate scientific discovery, fostering a future where technology and human ingenuity coalesce for rapid advancement.

My first @GnuOctave plot. ❤️

@skydotcs Can't wait for Gpt5! Just one ant emoji!

@0xluffy This was me at the very start of my MSc. It's still me, but I got 95/100 on my midterm. Nicotine is a performance enhancing drug.

day 98 of learning ML what the fuck is a matrix multiplication

@Michael_J_Black Academic work cannot be done by LLMs. They can only perform induction. Peer-reviewing is not a purely formal activity, therefore it cannot and will never be achievable probabilistically.

How do you rebut an AI-generated review? (This is not hypothetical -- I'm facing this now with ICCV reviews) When writing a rebuttal, I try to understand my reviewer. What do they know? What don't they know? What's annoying them? What do they really care about? But if a person submits an AI review, then I'm at a loss. Have they even read my paper? Have they read the AI review? What kind of person does this? Let's say I carefully rebut the points the AI brought up. What will the reviewer do with this information? They will have no context with which to understand it. Will they feed it to the AI? Why? My guess is that if they didn't bother to write a review themselves, they won't bother to read my rebuttal. All I know about my reviewer is that they are lazy and couldn't be bothered to read my paper, which I labored over, and give me their thoughts. I think the only proper response is to just not reply to their review at all. Or maybe I should give the review to an AI and ask it to write the rebuttal. Tit for tat. PS: Yes, I'm certain the review is AI generated. It is obvious. Bland and wordy with lots of **emphasis**. I ran it through an AI detector and it's rated as 100% AI.

@XorDev I disagree. Purely aesthetic design, maybe, but purely aesthetic design is basically meaningless anyway. Everything is just arranging nodes of meaning.

To be a good designer and a good programmer, is very hard. They are two separate skill domains with virtually no overlap

@nocontextmemes Drywallius

Not sure I've ever seen a head like that before, fantastic

Making the use of the blazing London sun today to get involved with some Datacamp.