27183

6 ms·

61.
by 27183 1mo ago
> The problem with optimizations is that you are competing in prioritization with other features. I think companies often over-indulge in features nobody wants, needs, or cares about. I quit my previous company because they were forcing
62.
by 27183 1mo ago
> What waste products do data centers output? Energy production consequences aside, they produce tons of e-waste
63.
by 27183 2mo ago
> Do you mean more earnings per share for corporations or real utility impacts on social systems (e.g., more new drugs) I mean real utility impacts. More reliable and efficient software, better features, etc. So far we've seen none
64.
by 27183 2mo ago
With http you can do it elegantly with a 429 and retry-after. 100% doing it at the edge is the way to go. The machinery which determines how long to delay a client's retry can benefit from knowledge of internal services' state, bu
65.
by 27183 2mo ago
Thank you for this candid answer. In the current climate of people breathlessly, hyperbolically jabbering about how AI is "revolutionizing everything" it's extremely refreshing to hear this honest, measured statement.
66.
by 27183 2mo ago
It's all about what you choose in the face of P. Do you choose C or do you choose A?
67.
by 27183 2mo ago
This can either be a relatively short time or an eternity, depending on the upstream consequences. If that means holding a connection open for 60000 milliseconds while waiting for some downstream rpc to go through its backoff ritual, that&#
68.
by 27183 2mo ago
It's easier to create a nice worst-case failure UX this way too. Rather than a page that sorta kinda loads but doesn't actually work at all you can unambiguously put up a "oops we're broken right now try again later&quot
69.
by 27183 2mo ago
OK I'll pick C and A! https://web.archive.org/web/20250128235041/https://codahale....
70.
by 27183 2mo ago
> ai code dev has enabled people to make tools for themselves that they didn't have before. ...that sounds pretty worthless tbh. The promise of AI wasn't that you can bang out your own little half baked duolingo clone. It was t
71.
by 27183 2mo ago
> That knowledge will be worthless 12-18 months from now when AI does everything better than you, including “understanding”. ...what? Seriously though. What grand insight into the arc of AI development do you have? Where are you gettin
72.
by 27183 2mo ago
Don't get me wrong, I think yours is a good example. It's important for people to have an example to reference when they want to resist but don't know how to do it, or who to contact, or where to start. But I don't belie
73.
by 27183 2mo ago
People who haven't spent much time around scientists or watched them work often have a mistaken idea of just how precise the work is. Even high precision work can be found later to be wrong! That's just the system working as des
74.
by 27183 2mo ago
just so long as you get at least one of them before they get you, you're fulfilling your life's purpose
75.
by 27183 2mo ago
> I make sure to probe these things during the interview process I absolutely plan to do so throughly when I begin interviewing later this year. But I've been fooled before, so it'll take extraordinary evidence to get my hopes
76.
by 27183 2mo ago
OpenAI, Anthropic, OpenRouter, Gemeni, etc also have another escape hatch: they can arbitrarily and potentially invisibly reduce your quality of service at their discretion. They can choose to route your request to an expensive model or a c
77.
by 27183 2mo ago
My pet theory is they backed themselves into a corner with Actions. That was a step change in computational scaling effort. It also fell immediately on the heels of the Microsoft acquisition. Something about these two factors seems to have
78.
by 27183 2mo ago
The resistance you're talking about only works if like 3% of the population take to the streets. Once you see police officers marching with protesters that's when they'll start to listen. That's only ever happened once.
79.
by 27183 2mo ago
well sure.. obviously. but only very few actually have any control over that and they're not incentivized to choose peace. the best the rest of us can do is vote in elections. it doesn't generally work out very well though.
80.
by 27183 2mo ago
The cost of fixing this will be everyone's problem. If we suddenly have to ramp up domestic production because diesel is >$10/gal it'll be a no holds barred unmitigated environmental catastrophe. We will be drilling and sp
81.
by 27183 2mo ago
If that's what they wanted they'd mandate that process. They want the deniability of having you rubber stamp the PR. That way they can blame you when it goes wrong. There may be high trust teams, but I've never seen one fir
82.
by 27183 2mo ago
A kilt is a convenient garment for this scenario
83.
by 27183 2mo ago
Try writing a regression test for your Github Action.. I'll wait. It's a pretty ridiculously half-baked feature. But a lot of these devtools are--it's arguably even worse in the observability vertical. You have all these metr
84.
by 27183 2mo ago
You can probably still write your own code to a standard of quality you're comfortable with, while conforming to the wishes of your employer. At least until you can't anymore, then you'll have to make a decision. But in this
85.
by 27183 2mo ago
If you hold up the line by flagging quality issues, you will be seen as the root cause of the problem. I say this as someone who has been there and done that--I've tried to do something about code quality issues affecting my team&#x2
86.
by 27183 2mo ago
Your bosses don't want you actually reviewing it anyway. They want you to approve the PR . Just give them what they want. I'm currently not working in tech, but I will again sometime within the next year or so, and I've been
87.
by 27183 2mo ago
Unfortunately we don't really know anything about OpenRouter or Stripe's numbers. They're both private companies. AFAIK there really isn't much information publicly available about how profitable or not OpenRouter actual
88.
by 27183 2mo ago
unless it's all just tulip mania, then they just spent $7B on nothing
89.
by 27183 2mo ago
Yes and for diesels there are options too, but fewer and as far as I'm aware none are open source or open hardware. If I'm going to bet on the longevity and maintainability of a computer control system all 3 of these things need t
90.
by 27183 2mo ago
I have a 1999 Benz E300, a 1995 Toyota FZJ80, and a 1998 Dodge Ram 2500. The Benz is by far the most complex, with full CAN network, drive-by-wire, and Bosch EDC fuel injection (no mechanical governor on the injection pump). The Toyota has

More ›