Does Sa… uh OpenAI still do stock clawbacks from employees who say negative things about the company after leaving?
Reflections on OpenAI
371–380 of 446 posts
Re: Reflections on OpenAI
#372Earlier quoted context omitted.
So this guy is filthy rich and yet decided to grind for 14 months with a newborn at home? I guess that's why he's filthy rich.
Way to go to keep the boring chores of the first months with the partner and join the fun when the little one starts to be more fun after a year. With all that cash, I'm sure they could buy a bunch of help for the partner too.
Yeah, or, let the partner have the easy period before they are mobile, and when they sleep half the day, and then join the fun when they can walk off into the craft supplies/pantry where sugar/flour/etc. are stored/the workshop with the power tools etc., and when they drop the naptime and instead start waking at 5am and asking you to play Roblox with them.
Either option is priceless :-)
Re: Reflections on OpenAI
#373[dead]
Re: Reflections on OpenAI
#374>Thanks to this bottoms-up culture, OpenAI is also very meritocratic. Historically, leaders in the company are promoted primarily based upon their ability to have good ideas and then execute upon them. Many leaders who were incredibly competent weren't very good at things like presenting at all-hands or political maneuvering. That matters less at OpenAI then it might at other companies. The best ideas do tend to win.…
I think in this structure people only think locally and they are not concerned with the overall mission of the company and do not actively think about morality of the mission or if they are following it.
In my experience, front-line and middle managers will penalize workers that stray from their explicit goals because they think something else more readily contributes to the company’s mission.
Re: Reflections on OpenAI
#375Engineers thinking they're building god is such a good marketing strategy. I can't overstate it. It's even difficult to be rational about it. I don't actually believe it's true, I think it's pure hype and LLMs won't even approximate AGI. But this idea is sort of half-immune to criticism or skepticism: you can always respond with "but what if it's true?". The stakes are so high that the potentially infinite payoff sno…
I know you're not being serious but building AGI as in something that thinks like a human, as proven possible by millions of humans wandering all over the place is very different from "building god".
Re: Reflections on OpenAI
#376Earlier quoted context omitted.
I am of the opinion that the greatest evils come from the most self-righteous.
That may very well be the case. But I think this is a distinct category of evil; the second one, in which you'll find most of the cigarette and gambling businesses, is that of evil caused by indifference . "Yes, I agree there are some downsides to our product and there are some people suffering because of that - but no one is forcing them to buy from us, they're people with agency and free will, they can act as adult…
It's not indifference, it's much more comically evil. Like, they're using software to identify gambling addicts on fixed incomes, to figure out how big retirees' social security checks are, and to ensure they lose the entire thing at the casino each week. They bonus out their marketing team for doing this successfully. They're using software to make sure that when a casino host's patron runs out of money and kills themselves, the casino host is not penalized but rewarded for a job well done.
At 8am every morning, the executives walk across the casino floor on their way to the board room, past the depressed people who have been there gambling by themselves the entire night, seeing their faces, then they go into a boardroom to strategize ways to get them those people to gamble even harder. They brag about it. It's absolute pure villainy.
Re: Reflections on OpenAI
#377What do people not like about Azure IAM? That's the one I'm most familiar with, and I've always thought it was decent, pretty vanilla.
When I go to AWS it looks similar except role assignments can't be scoped, so needs more duplication and maintenance. In that way Azure seems nicer. In everything else, it seems pretty equivalent.
But I see it catching flak occasionally on HN, so curious what others dislike.
Re: Reflections on OpenAI
#378Earlier quoted context omitted.
That part made me do a double take. I hope his child never learns they were being put second.
Many people are bad parents. Many are bad at their jobs. Many at bad at both. At least this guy is good at his job, and can provide very well for his family.
This is why the children of rich people are famously well adjusted... /s
Re: Reflections on OpenAI
#379Python monorepo is the biggest surprise in this whole article
Yeah legit interested in their toolchain. I tried Pants and had a bad experience. Bazel is too heavyweight imo and doesn't deal with a variety of dependencies well.
Re: Reflections on OpenAI
#380> What's funny about this is there are exactly three services that I would consider trustworthy: Azure Kubernetes Service, CosmosDB (Azure's document storage), and BlobStore. CosmosDB is trustworthy? Everyone I know that used CosmosDB ended up rewriting their code because of throttling.
It's pretty heavily used within Azure itself, and my old team didn't have any issues, even though we had a high volume service with stored procs, triggers, list indexes, etc. And it was cheaper and faster than the SQL Server instance it replaced (granted, the original Sql-based app had tons of big joins and wasn't designed to scale).
I know a few other teams when I was working elsewhere that had to migrate off Spanner due to throttling and random hiccups, though if Google uses it for zanzibar, they must not have that problem internally. Maybe all these companies increase throttling limits for first party use cases.
My current team uses dynamo, which has also given throttling issues, but generally only when they try to do things that it's not designed for (bulk updating a bunch of records in the same partition). Other than that, it seems reliable (incredibly consistent low latencies) and a bit cheaper than my experience with cosmos, though with fewer features.
They all seem to have their own pros and cons in my experience.