vanuatu

+1

I'd wager you get similar results if you gave people a version of Google search that purposely gave you bad results. Like, it's framed as an assistant / lookup tool - is it so surprising that people tend to trust it more? Especially since the participants are likely used to using full-powered models and the researchers give them a purposely gimped one (lol)

People are acting rationally when given AI tools to lookup information, their first consumer use case was as a super-powered Google Search

kind of a hilarious conspiracy theory

You are correct that LLMs are trained on existing proofs but hiring researchers to solve unsolved problems is just unrealistic, both in terms of how none of the mathematicians simply came out and took credit for their own discovery or exposed this, and how training sets are not easily memorized (rather, the meta techniques are learned).

OpenAI just has better training methods and techniques for pure math over Anthropic, it’s one of their biggest strengths

I play the game for a manufactured challenge

People in the real world all having god mode cheats is good, actually

LLMs are increasingly solving math problems on their own, producing entire correct proofs

I like Cal but his takes consistently miss the slope of improvement of the technology

This. the comments here talk as if YC is this secret cabal of upper elites when they're probably one of the least selective good accelerators

The top % of YC founders are legitimately quite talented, and high agency/entrepreneurial talent goes a long way in these companies, even if its not in deep infra or research

at the current pace, LLMs are the researchers

the end goal isnt to serve a chatbot to replace google or a coding agent to build crud

i dont think the comparison to traditional vc saas is very good

yes, in traditional venture you want cost per marginal user to decrease and leverage your platform at scale

but improving llms shifts the frontier of their capability and unlocks entirely new use cases. so far, every mega training run has resulted in a model that has paid itself off profitably fully loaded. perhaps the TAM of intelligence has no ceiling?

not to say that we shouldnt be investing in efficient models, but the efficiency comes after we create another mega shoggoth that we can make more efficient

"not enough people are emotionally prepared for if it's not a bubble"

- inference

- RLaaS (Tinker, or the more involved FDE motion a la Reflection / Applied Compute)

most companies will make the FDE role but not understand the value of the FDE org, which is to drive product strategy and function as R&D

Id say the main difference is FDEs post-engagement need to drive product strategy back into the platform (non trivial ask)

you typically see FDE-driven companies' products be 'assembly' driven and very deep into integration, as they figure out the optimal primitives that assemble into the shapes required to solve new customer problems

the main distinction i like to make is:

your FDEs shape your product strategy, and should be considered R&D. after making sure a customer deployment is successful (by any means necessary btw, even if it means building new systems outside of the product), the crucial next step is to drive the product improvement with PMs and core software engineers after contact with reality. this was a pretty radical idea from palantir in the era of saas

if you only do step 1 you're basically just solutions engineers / mckinsey, and if you only do step 2 with no customer learning to your product you don't improve your platform for all the other customers. the pain becomes the moat

There's a reason why this echelon of companies comp FDEs much, much more than services businesses is because you're trying to find engineering + product + customer facing in one (knew people making 200k+ 5 years ago as new grad FDEs, and the same flavour at the labs is 500k+ easy)

that being said the role has evolved a lot over the years, and depending on the company it could be indistinguishable from solutions eng, or sales eng, or even dev rel.

sadly with all the labs benchmaxxing I feel like you just have to try the model for a while to really evaluate how good it is, especially for each individual use case

100%, if someone from a no name school does well on the interview I’ll happily recommend them to be hired. However idk how HR filters resumes, and they likely use certain heuristics to try and minimize false positives

Do you need to do tier 1 and 2 work before tier 3?

If they are structurally different, and there’s a way to train people directly into tier 3, then it doesn’t seem unreasonable to automate t1 and t2 as from my experience the vast majority of the tickets are either simple or repeated workflows. Taking the idea to the limit, you’d automate all tiers, and have the ai escalate to the individual teams within the company for any truly meaningful edge cases

I feel sort of the same about SWE, which is much more complex, but juniors can ostensibly grow into seniors with AI

I joined a new company 6 months ago. I interviewed at 16 companies and got 5 offers from a mix of ai cos / big tech / trading firms

Background is SWE at an AI co that's in the news sometimes

It felt about the same in terms of grind effort from my last search in 2022. the main difference was ai companies cared a lot about your understanding of agentic systems and harness / context engineering, and had much more practical rounds with less leetcode (usually 1 medium). More legacy firms (finance / some big tech) still expected you to solve 3-4 leetcode medium/hards throughout the process

you can get 1% as a founding eng at seed, and its not uncommon for a 5 at 50 seed

dilution is also dependent on the opex, founder negotiating power, and growth of the company. There are startups raising monster rounds at <5% dilution a round

If you are an employee however and your co is raising highly diluted rounds with poor growth probably best to jump ship

For now!

Even if not, there's pricing pressure between chat, gemini, and claude. The products seem to be comparable for laypeople which is why OpenAI has been investing a ton in their memory feature to try to lock in users

It's unclear to me how this will play out because LLMs don't have the same network / platform effects as the other examples (Uber / Facebook), nor is there one dominant LLM that is overwhelmingly better than the competition for consumers (Google). There's overwhelming competition from the open source cheap models especially for the lower-mid intelligence use cases

not very often which is why its zero signal

in rare occasions it might go the other way around, like someone who has so much experience they dont need a pretty resume because their work speaks for itself

New grads are the biggest offenders of the resume slop

as someone who has done 100+ first round interviews for SWE, including new grad

zero signal: resume & cover letter. applicants will mass-apply with ai-tuned resumes that happen to perfectly match our listing

medium signal: top 15 school / top N internship experience / built something with paying users

highest signal: personal referrals

+1, referred candidates at my company perform way better in interviews and on the job compared to cold applicants

I'm very optimistic. I think most diseases being cured, extended lifespans, physical abundance, and zero poverty are within reach in our lifespan, due to technology.

I think humans will be about the same in terms of happiness, due to how quickly to acclimate to our situation. But they'll look back on us with shock at how we ever lived like this!

the article takes this into account with EV

obviously vol is not the end all be all but one of your main advantages as a startup employee is access to the insider info and being able to walk away

its not imagined, its already happening at scale in f500, the companies they work with to do this are going vertical

the govt helped lift many, many millions out of poverty since opening to the rest of the world?

its pretty easy to point your terminal agent to your giant pr and ask it to break it up into small prs

if youre being asked to rubberstamp prs thats a management skill issue

but we trust the stat that 16 pct of americans have a positive view?

an article on one person? cmon

it seems like Chinese people have much more faith in their govt than Western countries, and subsequently trust them more in distributing the benefits of AI (in aggregate ofc)