hellojimbo

I thought elon hates demis

Isn't that basically what browser base does. I've found the hardest part of browser use to be stealth first then client change management then browser comprehension (which gets better with every new model).

in the past, you genuinely needed other people all the time which made this so much easier

Anti nationalism was the liberal motto for the entirety of my lifetime.

i do

why not just start your own airline instead of buying one that is dysfunctional and losing money

For different brains, the answer has to be no because the images you see are a "neural net" construction and if that neural net differs then the "image" you see is different

PE has literally nothing to do with what Enron did which was accounting fraud + cashflow problems because they actually didnt make any money, in fact they lost tons of money and used future earnings in current reporting

we need the calories to feed a growing population

population doubles

we need the calories to feed a growing population

How does an autocratic leader lose an election, can Xi lose an election?

How can it be that groups pay for it long term when many of the successful apex predators exhibit interspecies murder and territorialism.

Just to use your own example https://en.wikipedia.org/wiki/Mapogo_lion_coalition

lol thats like 99% of planet earth, including the animals

Around the same revenue due to Anthropics strong enterprise strategy

The last time I typed ultrathink, i got a prompt saying that you no longer need to type ultrathink

Your argument is even worse lol. Obviously he's proposing that you can take your doctors note to the pharmacy and get what the doctor prescribed in addition to being allowed to self purchase behind the counter drugs.

When writing banal code, you can just ask it to write unit tests for certain conditions and it'll do a pretty good job. The cutting edge tools will correctly automatically run and iterate on the unit tests when they dont pass. You can even ask the agent to setup TDD.

You are either using it wrong or you are writing extremely niche code that has bad llm coverage

thank you for your helpful snark, the commentor couldnt have been asking just out of curiosity because he's evil

I can't tell if this thread is filled with bookie shills or (buy a mcNuke level)die hard libertarians or people who live under a rock and don't realize how pervasive sports gambling is.

Dont be dense, we are talking about "very few" in the context of draft kings which prints more money in 1 week than black market gambling has in years. The turnover in legal sports gambling in america is over 100b a year, this is larger than the gdp of most countries.

My takeaway from the episode was that its actually really easy to setup up visa, you just need to get the banks, vendors, and card issuers onboard, which should be easy if you're the government

They shouldnt be voting at all

The no tools part is impressive, with tools every model gets 100%

Theyre higher value than actual seniors because they have more domain knowledge, more neuroplastic, probably willing to work harder, etc

Yes

The real problem with hiring juniors is that tech does not have flexible pay. Realistically, hiring juniors is high risk high reward so they should naturally be on some kind of probational pay.

Hungarian jews are also extremely smart anecdotally.

Is this like the prince or art of war where we are supposed to draw some lesson from very specific critiques and extrapolate it to every scenario.

The only real point is number 5.

Huge vindication for what I have been saying all along: we need AI that integrates both neural networks and symbolic algorithms and representations

This is basically agents which is literally what everyone has been talking about for the past year lol.

(Importantly, the point of the Apple paper goal was to see how LRM’s unaided explore a space of solutions via reasoning and backtracking, not see how well it could use preexisting code retrieved from the web.

This is a false dichotomy. The thing that apple tested was dumb and dl'ing code from the internet is also dumb. What would've been interesting is, given the problem, would a reasoning agent know how to solve the problem with access to a coding env.

Do LLM’s conceptually understand Hanoi?

Yes and the paper didn't test for this. The paper basically tested the equivalent of, can a human do hanoi in their head.

I feel like what the author is advocating for is basically a neural net that can send instructions to an ALU/CPU, but I haven't seen anything promising that shows that its better than just giving an agent access to a terminal

Conventional "dev" wisdom says that LOC and PR count don't matter but I think its very easy to combine this data point with the nature of a dev's work to come up with a great heuristic on productivity.