Thelonious Oditamo

I ate lettuce today idgaf

@growing_daniel isn't the whole point that both panels should exist

Guilting people into thinking being ugly and disgusting is beautiful has gotta end

@alz_zyd_ are you scaling up the A/B testing for this experiment anytime soon?

if you need to learn real analysis don't even try to read Rudin just put all your pset questions into ChatGPT and talk with it until you understand the answers you'll do way better than the outdated and obsolete textbook readers

@i2cjak anthropic is obviously more alarmist and more polarizing wrt PR, but really unclear how that means they are evil.

there was a year when everyone thought OpenAI was the evil company but it turns out the company that professes righteousness may be the evil company. If only someone wrote a book about similar things

@BradleyKellard mans really wants to be a trillionaire, so he found a new way of calculating it

Jeff Bezos: "I've created $2.1 trillion of wealth for other people." "Somebody needs to make a list where they rank people by how much wealth they've created for other people." While talking about his net worth, he floated an idea that flips the whole concept of a "rich list" on its head. He did the math live, right there on stage. "Amazon's market cap is 2.3 trillion today. I own about 200 billion of it." "So if you take 2.3 trillion and subtract out the piece I kept for myself, then I've created something like $2.1 trillion of wealth for other people." That's not a typo. Trillion, with a T and it's not even his money. "That should put me pretty high on some kind of list." He didn't stop at himself, either. He immediately named who else deserves a spot at the top. "People like Jensen, Nvidia, he's going to be very high on that list." "That would be a pretty cool list. Somebody should do that list."

@alz_zyd_ During the pandemic my friend referred to movies as a "productivity hack" for precisely this reason

The default 2 hour length of a movie is poorly suited for the modern attention span. 30 mins seems like the sweet spot, 1 hour max

@samzliu I see what you're getting at, but think it's more complicated. In particular, the techno-optimism of SF and the Platonic idealism of "scientific minds" are often compatible, but also often at odds, and I would argue the former is significantly more dominant in the SF ethos.

I've figured out why SF culture is so fucked. It's not AI, constant schmoozing, or the gender ratio. It's the work itself. I met a journalist the other day that was fascinated about how people try to find humanity in San Francisco, a the city that seems dead-set on trying to remove humanity from our society. She was especially interested by a rule I have at my dinner parties: no AI, no start-ups, no talking about moving to New York. I felt an urge to defend the bay: How is this different from bankers, political staffers, intense military specialties, or professional athletes? All of whom work insane hours and often live in all consuming bubbles filled with non-stop competition, jargon, and monoculture? But I knew she was right somehow. And it has to do with what the process of becoming a "technical person" does to you. On the surface, technical work has the hallmarks of flow. The mythical state of consciousness ascribed to only the best of us: artists, musicians, athletes, monks meditating, and the commuter driving home who has learned to stop worrying and love the ride. Long periods of time pass without conscious awareness. Focus that has its own inertia. Collapsing in of the world to only what matters right now. However, I often find that deep technical problem solving is the junk food of flow states. It maybe feels good in the moment but it leaves you feeling bad about yourself when the spell breaks. Similar to wasting a night away binge watching tv or playing video games until the sunrises. It’s a state where compulsion for what is next matters more than present joy. Not all technical problem solving feels like this, of course. There are moments of pure unadulterated beauty that I cannot experience in any other way. Moments where the struggle to the top of the mountain seems worth it because of the view. Maxwell, Noether, Euler. Giants whose first ascents have carved their names into history and whose paths have shown us the righteous way. Then there are the moments of overwhelming joy, relief, and exuberance that comes all at once. When you finally fix a bug. When the product is launched and the rover lands safely. The feeling that you have built something. Something complex and it works. And for me, those moments are what keeps me coming back. But it’s self-delusion to pretend that all or indeed most technical work is like this. Most of it looks like frustration. confusion. compulsion. suffering. Those that learned to code pre-LLMs remember the pain of debugging. Chasing down stack traces for hours only to find a stupid one letter mistake the cause of all your agony. The panic at 6 am, mere hours before a problem set is due, with 3 more questions still left undone. The utter confusion and hopelessness of concepts not quite clicking together. Or, once it clicks, it becoming so obvious that spelling it out in code or writing becomes an exercise in tedium. It's also not quite like a runner’s high either. There is no beauty found in this suffering. There’s no poetry in finding a bug that shouldn’t have been there in the first place. There’s no value in being confused. Indeed, the ones lauded here seemingly don’t have this problem. The prodigies for whom everything comes naturally. Perhaps this is why many people in deep technical subjects like physics, math, and theoretical computer science develop a hint of arrogance. If you haven’t suffered like we have, who are we to take you seriously? It is natural for the language of your craft to become the language of your life. And there’s a way in which a mathematical framework naturally lends itself to describing life. As has been famously observed, it is unreasonably effective. The physical world becomes reduced to forces. The world of human decisions to utility functions and Nash equilibria. This reductionism can be quite jarring for those that are non-technical. I remember accidentally offending a former partner by using the phrase “signal-to-noise” ratio, forgetting the connotations of the word "noise" in everyday speech. Just as mathematical thinking reduces the world to mechanical logical processes, the act of doing math reduces your brain to a machine. This is the crux of the problem. Unlike other forms of flow like music or sports, doing math inherently forces a type of dissociation. An abstraction of your mind from your body. To simulate an entire reality in your mind, you must remove yourself from this reality. Unlike prose or verse, the reality of mathematics and machine is devoid of texture. Almost poetically, everything here exists solely because someone willed it into existence. But in doing so, every object is perfectly smooth, platonic and pure. A mathematical definition is perfectly transferable without identity. It is the same to me as it is to you. Then, when you are transported back to the physical, embodied reality of the present, the whiplash catches up. All the sensory input, negative feelings you've been avoiding, and real life to be lived appear all at once and become overwhelming. It might be easier to just stay in the fantasy world where everything is simple and makes sense. Moreover, unlike other forms of disassociation which waste time or money (e.g. gambling, doomscrolling), a technical mind is rewarded extremely handsomely by society. For those that grew up in families that were emotionally repressed and valued success, it would seem worth it. Same for those that maybe have found solace in the simplicity of machines compared to the messiness of people. The cost though, is a sense of who you are.

@deanwball ALL CAPS when you spell the man name

I have had that mf doom song from the anthropic “keep thinking” ad stuck in my head all day, great ad, easily my favorite ad for computation since the apple louis armstrong facetime ad

people are very mad about the anthropic ad (probably rightfully so), but i also don't remember the last time this many people were talking about an ad since the anthropic super bowl one lmao

@BenSManning they gotta replace the flurry of professors "on leave at [insert name]", because those ones are never coming back

Berkeley is absolutely raking in the Academia AI talent... sheesh!

@emollick are there any good thinkpieces that should be given more attention? Most of what I see on twitter (and even substack, sadly) is extremely drab

Between having AI as a subject for infinite speculation and using AI as a writer to produce such speculation at scale, it is the Golden Age of thinkpieces.

@pfau are they rooting for Messi or against the Brits

I literally live in London and half the people in my office are rooting for Argentina.

the reason messi vs. ronaldo was ever a debate is bc of the champions league success in the late 2010s. but then madrid won 2 UCL's without him, and everyone realized it was just the madrid dna carrying. 2021 onward has just been icing on the cake

@FakePsyho tone and delivery are often everything, for better or worse

twitter today

@littmath Gave it an open problem and it's been thinking for over 30 hours. All of the intermediate messages are complete gibberish (don't even sound like math), but I'm too bewildered to not let it organically run. If it responds in < 1 hr, I've been quite impressed.

Overall I think 5.6 Sol Pro/Ultra etc. seems to be a substantial step up from 5.5 for math. That said, common interaction pattern is: I ask a question. It thinks for ~100+ minutes and returns a largely inscrutable response. I ask it to explain. It thinks for 20 minutes and says:

@dylan522p datacentral park

I would like to see them TRY STOP THIS

@WillManidis ragebait a little too obvious

it feels insane to be using images of the final resting place of those that gave their life for our country when anthropic won't even give their models to defend those very same warfighters

@frontier_foid many of the top olympiad/imo ppl ofc are also finding themselves in ai now, but more as ICs and not on the founder level

idk why people refer to their backgrounds as competitive math when most of the folks here were much better at ioi than imo (im actually not sure if any of them even went to the imo?). in fact, the best imo participants of this era haven't done much of note in ai!

@tolstoybb defecade

the diarrheal decade

@StatisticUrban china?

Getting to be quite laborious, no? If we want to go back to single-country hosts at any point, only a few countries on Earth could hold a tournament this large using FIFA's preferred stadium size.