gfourfour

This seems like an extremely petty fight

“Taste” is something that’s developed through repetitive exposure to differentiated items in a particular set, combined with extremely high abstract analytical abilities, and that’s something completely different from having marketing or personal branding skills.

I think you’re right that that’s the evaluation happening, but it’s totally misguided. If you’re indexing for differentiating levels of taste I would be very wary of empty vessel young influencers. Taste is built over years and years and imo requires a certain disdain for the crowd. Look at Linus Torvalds as a pinnacle of taste in code for example.

I think wrapper companies could do very well if AI does end up becoming having an impact near on the order of the internet or smartphones, which it may never.

A product that incorporates a foundation model could still require solving a hard engineering problem separately from the training and provision of said model. Someone who could, say, provide a reliable filter for copyrighted content from an LLM is providing real value and has a tangible moat.

Acting like Elon Musk is cringe for a lot of reasons but I don’t think it’s that out of pocket for a startup founder to admire him, even if all he did was SpaceX he would have an incredible legacy and story

I switched to concerta and it works much better. Lowering my dose of adderall just made it less effective.

Adderall is great at making me produce, but horrible at true “focus.” It makes everything other than pure output feel like a waste of time. In the long run, I found this to be a pretty detrimental way to operate. Good for getting something across the finish line but since I’ve stopped taking it I’ve been able to go a lot deeper into stuff and plug away at long term projects rather than just checking off as many “tasks” as possible.

Diminishing returns on that though, and it comes at a cost to your physical and emotional health.

The mountains are awesome but idk how many people are moving to Seattle for the weather haha

Yeah UW is great but one university isn’t enough to sustain a world class talent pipeline

Seattle/Washington has no elite or quasi-elite universities. That’s probably the #1 reason.

Compare it to Boston which is a similarly sized city.

Boston/New England has Harvard, MIT, BC, Tufts, Dartmouth, Amherst, Williams, Middlebury, Brown, etc. It’s the most educated region of the world with huge pool of founders and talent. Even though it’s not a tech focused place the way Seattle is it’s a better place to find young talented people willing to work for or start a new company.

I can’t imagine that DC is a better place to have a startup than Seattle

Even if credentials matter more than ideas there’s more than enough credentialed people to go around and soak up funding

Plus, tech isn’t the only startup dimension. Boston has a ton of medical stuff too.

Being outside for an hour during sunset is the best sleep medicine. Don’t even have to go camping.

General knowledge is 90% from interests and consumption outside of school

“Solving for colour constancy using a constrained dichromatic reflection model” is a good one to check out

Wow, your second paragraph hit the nail on the head about the effect that any sort of ML/AI work has on me. Tweaking hyperparameters/prompts/features and running it “one more time” is like sitting down at a slot machine for me. Can burn a day without any real work done iterating towards some unreachable perfect run.

How does an iPhone contribute to censorship?

I think this is a really underexplored use case for LLMs. LLM embeddings are really good at encoding rich semantic information that’s easy to query and hack around with in a variety of ways. Retrieving primary sources that correspond to one or many thematic dimensions is one such case for embeddings, but most applications that do this portray it as a driver of RAG chatbots, when it could be an end in and of itself.

I have an app that does your book recommending idea but with Wikipedia articles. I am trying to release it soon, once I get past my perfectionism, if anyone is interested. Expanding to non Wikipedia sources is an eventual goal.

I basically never want to read chatbot output for pleasure. I want to read primary sources.

I had a similar experience where I gained a strong sense of how ancient humans and modern humans are one and the same.

It was individualistic in that the sense came from my own embodied experience of being in the world, and I was thinking about myself, but it was an insight about contextualizing myself in a much larger thing.

It made me feel special, personally, for being a member of such an incredible group, but anti-individualistic because I extended this specialness to everyone.

That’s not why.

Credit scores don’t discourage competition. Competition can still exist within brackets.

The reason rates don’t differ much is because most of the rate is determined by factors outside of the lenders control. There’s a pretty small margin for them to compete on price with each other with. That’s why rates on stuff like mortgages are basically commoditized.

I still use Reddit in the way you’re suggesting, which is when I occasionally have a specific use.

What I meant was that I don’t visit Reddit by instinct anymore when I’m bored - even “good” subreddits. The general user base is not high quality. Even for topics I’m interested in, I really can’t stand the Reddit vibe that inevitably seeps through. It’s incredibly predictable.

A farmers market has a narrower selection than a supermarket, but you’d be healthier in the long run only shopping at the farmers market than going to the grocery store and being bombarded by processed garbage, even if you try to stick to the good aisles.

The problem with Reddit is that the worst of Reddit inevitably pollutes the best of Reddit. Better for your mental diet to avoid altogether.

I “de-reddited” with HN.

The philosophy students were right, some of the highest mid career earnings of any major. It’s harder to get your foot in the door without specialized knowledge - but careers arent determined by your first entry level job.

I’ve seen boards hinder nonprofits from executing on the mission plenty of times, usually out of ego or general fecklessness.

I wouldn’t assume that boards care about the mission more than the employees. Employees are the ones signing up to implement it, board members are often just in it for status and are total dilettantes.

I think in this case there’s a lot of signs pointing to the board feeling that their status was threatened more than anything. If their reasoning is that releasing ChatGPT without their permission was “unsafe,” give me a break.

I was going to use ec2 and s3, should I look at digital ocean or hetzner instead?

To be honest, my bias is against boards, especially non-profit boards who have no skin in the game and often an idealistic view that clashes with operational realities.

The case of Altman seems to be a “where there’s smoke” thing to an extent, but I also am not inclined to trust the board over the employees who do seem to like him.

Insofar as he steered OAI to be a consumer product company rather than a research institution, that’s an acceptable outcome to me. The board had a problem with that, fought him, and lost.

No haha just a rats nest of a bunch of notebooks

mwparserfromhell can parse the text content without including tables

Still, $5000 is kind of insane. That makes no sense to me.