qoez

The west? Plenty of european schools are extremely tough. I think this is mainly an american phenomenon.

ML when connected to business applications is one of those things that sound complicated but is in practice just a few lines of code usually. I'd argue that anything frontend is way harder to code up than a pretrained TF lite model, despite the general idea being the opposite.

Doesn't take much ingenuity to code up a basic pretrained TF lite model to analyze an image.

I mostly agree. Sports are mostly very well priced (live sports not so much). Many new markets that are just invented by polymarket/kalshi though SIG are basically on an even playing field with you in terms of coming up with a way to price it (and if it's a niche low liquidity market they probably don't pay much attention to it anyway so you can easily come in and price it with a large edge). Personally lots of my money is probably coming from SIG itself in fact given that I mostly take liquidity that are placed at poor prices where I've spent many weeks on trying to figure a more informed way to price it.

That's obviously true, but at least retail are totally allowed to place maker orders and act as the 'sportsbook' themselves. Taker orders are usually (but far from always) -ev but again at least there's the option to be a maker. Many of the best opportunities is to take liquidty from HFT firms with poor prices as well (there's too many markets for them to price it all correctly and you have an advantage by being allowed infinite time to think about how to price a given market).

As someone who makes their living doing PMs this is far from the reality. There's lots of markets that are beatable without a latency edge. Also anyone can apply to be co-located with their servers if you do KYC.

I get the sense that they're trying to manufacture the AI narratives of the 80s so they get to control the story when it happens. My semi conspiracy theory is that anthropics biology wet lab is also going to lead to some kind of AI manufactured bio event (which anthropic said they predicted will happen in a few years).

Everyone is talking about the price but $300 does not seem that much for what this offers imo.

To be fair we've gone from barely capable at writing code to what we have today in barely 3 years. I wouldn't put money on the counter bet that they won't get beyond this hump.

There's lots of good interviews from that series. I enjoyed the link's awakening interview as well

Watching the documentary based on this made me realize after years hearing about this that it's sort of incomprehensible word slop that I doubt many of the people recommending it got anything from at all https://www.youtube.com/playlist?list=PL2SCoeb8QMtWWD5_7tiYB....

Feels like some kind of sentence to make an example of someone to scare off other people ala going after grandmas for napster. There's way too much of this going on on instagram and tiktok etc etc already. They went after this guy but is letting many people get away with it.

Without a thriving mathematical community to point out these things it would have stayed broken. With automated math that community as tao pointed out is at risk.

Seems silly not to have predicted that 10 years ago. I feel like it's long been obvious that smarter models being available will mean way easier cheap synthetic data and access to tools that will speed up competitors as well as consumers.

The remaining 5% and their adapted mutation will eventually repopulate the local population.

To be fair we have had UI slop way before AI in the sense of lazily coded react slop.

I always assumed this was an america-only drug that no one actually does or at least not to the excess that it's claimed by the publications using it for the normal fear journalism purposes.

Fundamentally I feel like coders just doesn't even need to be that smart anymore given AI assistance. This place ten years ago used to be filled with some of the most interesting comments/takes around for that reason.

The non overreacting take is probably that young people will still have plenty of interaction with parents and their peers etc to develop critical thinking as a guard against deception and whatever else. Definitely some low single digit people without a social support system around them will develop AI psychosis though.

That's what I like most about codex, it has the 'pragmatic' tone mode which basically makes it tone-less. If claude's tone was the only one available I would seriously consider not even using LLMs or switching to a career that didn't require it because I literally can't stand interacting with it's personality for many hours each day.

This article might be good but I always get scared when they use AI art for the header bc I wont know if I'm reading AI slop until i've invested a good amount of time into the article so I just avoid it all together.

Seems like another case of LLMs taking credit for their training data without giving proper attribution, similar to the navier stokes case.

America is 'scaling' their low-trust society into the rest of the world and industrial levels

Collaborative research is a dual purpose existensial safety hazard, it's better to all results locked up away from public use for the public's own benefit.

This is basically a version of rage bait that I find most productive trying to stay away from reading

"Did Tristan opt out of model training for his ChatGPT and Codex sessions? If the answer is no, then this seems fair game"

That's assuming they actually honor this which I'm highly sceptical of. Especially for internal frontier models

"That’s part of its appeal — it’s a one-of-a-kind reality-TV-as-hangout-movie psychodrama circus"

This is fable level unreadable to me

Maybe if I develop really strong anxiety about this, it's gonna solve the problem

Makes me wonder if openai is secretly querying claude for some prompts and that this is why it's down due to spacex's data centers being down.