Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bsaul
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
49 ms
·
661.
▲
by
bsaul
4y ago
Yeap that's my point. Tracking isn't the issue, tracking by google is.
662.
▲
by
bsaul
4y ago
Such a typical example of someone completely rational, forgetting humans aren’t only about logic. People have emotions, doubts, suspicions, trust issues, affect, all of which aren’t easily quantifiable. The reason people don’t want telemetr
663.
▲
by
bsaul
4y ago
i don't think joelonsoftware points apply here. - Their product is still relatively new (7 months of coding isn't that much) and not even released, - they picked a technology specifically designed for the problems they've i
664.
▲
by
bsaul
4y ago
Wonder how long it'll take until SF switches side to either republicans or independent. Maybe once even the super rich leave, and all is left to vote is the crackheads.
665.
▲
by
bsaul
4y ago
This is a bit sad that someone like woz who was at the heart of the computer revolution, on the engineering side, was unable to anticipate how self driving wasn’t anywhere near the corner at the musk sold it. It’s a bit scary as well for an
666.
▲
by
bsaul
4y ago
My experience was more 50% bullshit 50% fact. I did however explicitly forbid members of my team at work to use its code answers for subjects they weren't already experts in. However I'm not advocating using its answers directly,
667.
▲
by
bsaul
4y ago
The fact that i can use this tool as a source of inspiration, or a first opinion on any kind of problem on earth is totally incredible. Now whenever i'm stuck on a problem, chatgpt has become an option . And this happens in the arti
668.
▲
by
bsaul
4y ago
I think this time is the good one. ChatGPT has reached a level where we finally can think of building actually useful products on top of "AI". Note that nobody is pretending that ChatGPT is "true" intelligence (whatever
669.
▲
by
bsaul
4y ago
how robust is nats persistence layer ? strenght of kafka relies on its proven ability to maintain a high load while making sure you won't loose the events you're streaming.
670.
▲
by
bsaul
4y ago
With this kind of reasoning there would be no way to disrupt a first actor in a new field.
671.
▲
by
bsaul
4y ago
except it would be perfectly acceptable for a comedian on stage to say fuck or cunt, provided the context makes it funny.
672.
▲
by
bsaul
4y ago
well then maybe that's why. Your brain is just telling you it needs to breath a little.
673.
▲
by
bsaul
4y ago
Sometimes problems in your personnal life affect your work. Make sure there isn't something that bothers you in that space. Also, when was the last time you had a 1 week holidays ? US companies grant very very few holidays, it's n
674.
▲
by
bsaul
4y ago
« AI doesn’t get an n-word pass » Even a black AI ? See how the situation is turning ridiculous ? Of course an AI should be able to say any word as long as the context makes it clear that it’s not actually insulting anyone. Racism, phobia o
675.
▲
by
bsaul
4y ago
As many people here noticed, the bit has actually nothing to do with transphobia, or gay, but it's about a comedian bombing a show with horrible jokes. What's interesting in this case, is that there can be no "background chec
676.
▲
by
bsaul
4y ago
i know the "pass sync.mutex by value" error was diagnosed by my linter, but are there any plan to fix all those issues in the go compiler itself ? Thoses are really bad footguns.
677.
▲
by
bsaul
4y ago
Everytime i read stephen wolfram, i can never decide if he’s a genius on the brink of a groundbreaking discovery (emphasis on « on the brink », because i don’t think he has released anything proved to be groundbreaking yet ), or just a for
678.
▲
by
bsaul
4y ago
my point was that analytics on the web is already quite known. Everyone tracks your every move, tiktok won't be able to do much worse.
679.
▲
by
bsaul
4y ago
it's going to be way harder to explain how tiktok is able to collect more personnal data than the regular ecommerce website.
680.
▲
by
bsaul
4y ago
You're confounding the problem of feeding a system with incorrect inputs ( calculator ), with a system that, provided the input are correct, will throw garbage totally at random. This is a very different kind of system with very differ
681.
▲
by
bsaul
4y ago
damn, just when i thought the "1 new js framework a day" race had calmed down, i'm reading your comment and realize it hasn't one bit :)))
682.
▲
by
bsaul
4y ago
Continuously learning chatgpt would probably be absolutely insane indeed. Yet it remains that chatgpt has no notion of « truth » or information quality. Whereas google has built a whole system of ranking reliable source of truth for every p
683.
▲
by
bsaul
4y ago
Please make sure you record the talks and put them on youtube , then repost on HN when they’re out. I’m sure lots of people here will be highly interested in what those people have to say !
684.
▲
by
bsaul
4y ago
Some people really put all their effort trying to make the worst conspiracy theorists nightmare come true. It's as if they're doing it on purpose, it's just crazy..
685.
▲
by
bsaul
4y ago
I have the feeling we really don't know what it's good at for certain. Due to its highly statistical nature, and completely opaque data structure, we can only observe some behaviors, empirically. But it's a total black box.
686.
▲
by
bsaul
4y ago
From an aestethic pov, swift is taking the exact opposite direction than go: it keeps piling on more language features, with much less concerns on keeping the language simple. As a developer currently working in both, it feels like jumping
687.
▲
by
bsaul
4y ago
this post makes me realize why swift is starting to look more and more bloated from an end-user pov. It was designed to be right from the beginning. instead of carefully backing off complex problems and slashing features to keep the design
688.
▲
by
bsaul
4y ago
Do you think real people will get confused and think the only way to write hello world in bog is using a variable ? Or do you think most people will automatically adjust and think "hello world in itself doesn't show much, so the a
689.
▲
by
bsaul
4y ago
Isn't it well known by now that chatgpt isn't reliable enough to build anything really serious on top of ?
690.
▲
by
bsaul
4y ago
From a tech perspective, e2e adds a lot of constraints on reliability and message ordering which makes it hard to distribute (think « exactly once » vs « at least once » kind of problem, albeit a little different). It also adds a lot of UX
More ›