Eridrus

25 ms·

391.
by Eridrus 3y ago
People keep saying that training is tricky, which is true, you need people with experience, but they're not so rare that this will present a moat forever since you can very much hire the people who know how to do this. My experience wi
392.
by Eridrus 3y ago
From talking to senior folks at tech companies, there's a real perception that, modulo the first few months of covid where there was not much to do hut work, productivity has suffered at their companies. In other circumstances (eg MOOC
393.
by Eridrus 3y ago
> There was nothing ready or thought through. This was not my view being somewhat close to the technologies there. Lamda did exist internally for several years, with pretty good demos. It was clearly not GPT-4, but there was a lot of exe
394.
by Eridrus 3y ago
Academic studies of social media are often very hampered by tooling and data access and studying a moving target. It's hard to even know if the methodology of the paper you cited (analyzing comment trajectories) is a good one, given YT
395.
by Eridrus 3y ago
I don't know YT's monetization history, but longform video is incredibly easy to monetize because advertisers are willing to pay much more for their content being there. They get some edge from all the tech they have built for mat
396.
by Eridrus 3y ago
> Regulatory and PR risks are similarly grave. For example, Google couldn't have pulled off something like TikTok without all kinds of regulators jumping at their throats right away. They had to wait for ByteDance to clear the way a
397.
by Eridrus 3y ago
Does protobuf or capnproto have good tooling support out in the world? Proto was very convenient at Google since everything was proto, your debug tools were proto, all your data pipeline tools spoke proto, you had command line utilities for
398.
by Eridrus 3y ago
Would the apprenticeship program not have the same problem as HR regularly does?
399.
by Eridrus 3y ago
Not to deny that founders are in a privileged position, but they are still beholden to other investors and have clear incentives (and legal obligations) to not screw over current employees. Even when founders control the board, the company
400.
by Eridrus 3y ago
I agree this is a natural thing to try. Given I have not seen anyone report success with this, I lean towards thinking this doesn't actually work reliably (or the entire community is asleep at the wheel).
401.
by Eridrus 3y ago
The Universal Interface currently has a huge problem in the form of "prompt injection" where currently we have no principled way to separate instructions and data, e.g. https://embracethered.com/blog/posts&#x2
402.
by Eridrus 3y ago
I am unaware of any situations where employees have been diluted out but founders have not. There are cases where everyone gets diluted out because the company is doing poorly and needs to be recapitalized, but in those situation everyone&#
403.
by Eridrus 3y ago
I didn't hear about this, nor could I find anything specific on Google. I did find that apparently some companies have pretty aggressive clawback clauses, which is a huge yikes from me. Anyway, I think if you ever do talk to a mid stag
404.
by Eridrus 3y ago
You don't need an advanced degree to work on these technologies at big tech either.
405.
by Eridrus 3y ago
How do you feel about options that can be exercised for 7+ years after you leave? Does that change your calculus at all.
406.
by Eridrus 3y ago
I think there is a clear mismatch between how founders and employees value equity. The advice on the internet is to always ignore the equity component, but founders are definitionally there because they think the thing they are building wil
407.
by Eridrus 3y ago
> Unfortunately, they were surprisingly effective at the thing I care about which is collecting millions of people's personal information without their knowledge or consent. I think this is the core disagreement, I do not see this i
408.
by Eridrus 3y ago
I have two thoughts on Cambridge Analytica: a) Influencing elections is not a harm, even if you disagree with the political ends. b) They were notoriously ineffective, which further solidifies this into the non-story bucket
409.
by Eridrus 3y ago
Go on, follow that train of thought.
410.
by Eridrus 3y ago
Do you have any evidence of an actual harm being inflicted on a meaningful amount of people that would be sufficient to shutdown such a large amount of economic activity?
411.
by Eridrus 3y ago
It is not a core cloud service, it is a service for a niche industry that never met the hype it had and probably didn't get many customers. I don't think discontinuing cloud iot provides any indication that compute engine is going
412.
by Eridrus 3y ago
If your company is that big and unwieldy that a tiny chance of needing to do a migration in 12 months is unacceptable, you can stick to the key services which offer 36 months of notice. For all I know at that point you can negotiate your ow
413.
by Eridrus 3y ago
Lol, ok man, sounds like you're real good at making rational choices. I'm sure there's exactly zero chance AWS or Azure will ever have any problems.
414.
by Eridrus 3y ago
Sure, I used to work at Google, I used to work at Microsoft, some of my best friends worked at Amazon, I own stock in all of them. Are you interested in making a bet on the chance GCP will be discontinued?
415.
by Eridrus 3y ago
A 5% chance of needing to do a migration in 10 years is very small in comparison to other technical risks. And lets be clear, this is not a catastrophic risk; your contract with GCP guarantees they will not discontinue services without 12 m
416.
by Eridrus 3y ago
If you want to give me 10:1 odds GCP is not going to be around in 10 years, I will bet you any amount of money you are willing to name lol.
417.
by Eridrus 3y ago
I'm tired of reading the same thing over and over where people prognosticate based on poor readings of events and I think people should put there money where their mouth is. I really do not think the chance of GCP discontinuing a core
418.
by Eridrus 3y ago
At what odds are you willing to make a bet on this? Do you think there's a 20% chance it's gone in 5 years? Do you think it's 5%? Do you think it's 30%? People keep saying things will be shut down and I think they should
419.
by Eridrus 3y ago
I'm not going to read yegge's rants. Name the odds and condition you think gcp will violate.
420.
by Eridrus 3y ago
What odds would you like to bet that GCP is going to be gone in 10 years?

More ›