Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sheeshkebab
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
61.
▲
by
sheeshkebab
2y ago
Swift is useless outside of macOS/iOS dev… the thing doesn’t even have namespacing. I don’t know for what reason someone would use it outside for Apple environment.
62.
▲
by
sheeshkebab
2y ago
Maybe Musk will show to all how it’s done starting with robotaxis, however looking at current state I wouldn’t bet on it, at least not anytime in near future. Buying more gpus will definitely not be the thing that solves this.
63.
▲
by
sheeshkebab
2y ago
The only thing AI needs to do to demonstrate its value is replacing jobs at scale. Until it happens, it’s all capex and opex bs and hobby projects.
64.
▲
by
sheeshkebab
2y ago
Show me a test and I’ll show you a neural network that passes it… used to be an saying.
65.
▲
by
sheeshkebab
2y ago
In your head. You just make some new weird shit up and try it to see if it works any better than the previous shit you made up.
66.
▲
by
sheeshkebab
2y ago
epic… and not a single of these “experts” likely can solve even a basic goat problem https://x.com/svpino/status/1790624957380342151
67.
▲
by
sheeshkebab
2y ago
Talk to Microsoft
68.
▲
by
sheeshkebab
2y ago
> The next revolution is going to be when people realize what we have already Enlighten us
69.
▲
by
sheeshkebab
3y ago
The article is written with errors and sounds like it was created by a bot, complemented with some fake quotes… I guess the battery is ok or is the point here to push BYD cars?
70.
▲
by
sheeshkebab
3y ago
it seems you can add fluff to existing projects with this round of AI but replacing labor that goes into dealing with the core of the tasks not so much. I’m not sure gen ai even helps with increasing productivity yet - since tons of existin
71.
▲
by
sheeshkebab
3y ago
What labor did the current AI save yet? All I see so far is billions or even trillions in capex for stuff that barely works…
72.
▲
by
sheeshkebab
3y ago
And they should
73.
▲
by
sheeshkebab
3y ago
Why then all the talk about AGI when fundamentals don’t even allow for it to emerge.
74.
▲
by
sheeshkebab
3y ago
It’s indeed odd that current dnn’s require massive amount of energy to retrain and lack any kind of practical continuous adaptation and learning.
75.
▲
by
sheeshkebab
3y ago
Jon should have stuck to comedy… there is a lot of angry news shows as it is.
76.
▲
by
sheeshkebab
3y ago
It feels like math issue is solvable, so are an all the other more precise engineering type things (even basic calculators can do all that). Yes, it’s early and but it seems like llms are on the right path this time, and what the article de
77.
▲
by
sheeshkebab
3y ago
Oh yeah? I guess you never day traded..
78.
▲
by
sheeshkebab
3y ago
Fedramp? High?
79.
▲
by
sheeshkebab
3y ago
No one wants to talk to a chatbot.
80.
▲
by
sheeshkebab
3y ago
Elon is a genius
81.
▲
by
sheeshkebab
3y ago
Get rid of prescription requirement too - who needs to see a doc for extra $500 when I can go and pick my own med, and possibly better one.
82.
▲
by
sheeshkebab
3y ago
Quite some lofty goals for $400k investment… supplanting silicon chips no less
83.
▲
by
sheeshkebab
3y ago
Amen, local first should be the default for anything that sucks all my data. Although until these things can do my laundry none of them deserve any of my compute time either.
84.
▲
by
sheeshkebab
3y ago
20+ years later, and billions of lines or python code later, the GIL discussion is still there. Kind of an amazing mess of a problem… mainly shows the priority of python being a ducktape type solution from the start.
85.
▲
by
sheeshkebab
3y ago
But can it play Doom?
86.
▲
by
sheeshkebab
3y ago
The important part about current ai craze is that it’s the first realized approach that somewhat resembles all the theories, fantasies, and bull shit about it. Can it eventually lead to superintelligence? it seems like it’s stepping in the
87.
▲
by
sheeshkebab
3y ago
Anytime I ask these things something (bard, gpt etc), 33% of the answer is genius, 33% misleading garbage, 33% filler stuff that’s neither here or there The problem is distinguishing between these parts requires me to be be an expert in t
88.
▲
by
sheeshkebab
3y ago
DX short funds spreading fud?
89.
▲
by
sheeshkebab
3y ago
Oh, hiring 1099 consultants now looking rather attractive
90.
▲
by
sheeshkebab
4y ago
Public domain applies to uncopyrightable works, among other things (including previously copyrighted works). In this case models are uncopyrightable, and I think FB (or any of these newfangled ai cos) would have interesting time proving oth
More ›