Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
BobbyJo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
91.
▲
by
BobbyJo
5mo ago
Having many semantic options for error usage is functionally the same as having many error types, except worse.
92.
▲
by
BobbyJo
5mo ago
They are there for attribution and, depending on the license, preventing people from making money off something that is supposed to be free. Attribution from an LLMs output is nearly as infeasible as attribution of where I learned the words
93.
▲
by
BobbyJo
5mo ago
That would mean that the A.I. is no longer "plagiarizing" in your view?
94.
▲
by
BobbyJo
5mo ago
Is there anyway to build intelligence that doesn't meet the definition of plagiarism you are using here? I remember when IP laws were looked at like a form of oppression in the tech community...
95.
▲
by
BobbyJo
5mo ago
It's hyper palatable food in the form of conversation. I see society treating it the same way eventually, at least along this one axis of interaction.
96.
▲
by
BobbyJo
5mo ago
> So you're looking at about a 10 year lag from best consumer GPUs to a GPU with similar performance to a modern phone. Two competing viewpoints to this: 1) It is getting harder to make the same performance gains, so maybe that 10 y
97.
▲
by
BobbyJo
5mo ago
I would go even farther and say that static types are a tool designed specifically for a code reader . When you're writing the code, you know what the types are, as you literally just created/wired/whatever them. Static ty
98.
▲
by
BobbyJo
5mo ago
LLMs do fantastic when you do the architecting for them. Don't let them make system decisions, and you'll have a great time.
99.
▲
by
BobbyJo
5mo ago
OP was saying not to join because you'll have a shitty time, not because the products aren't inspirational enough.
100.
▲
by
BobbyJo
5mo ago
If AI became a forcing function for cross industry semantic consistency in public tooling, I would be so happy.
101.
▲
by
BobbyJo
5mo ago
Computers really aren't that valuable to the average person who already has a smart phone. For everyone else, many probably have a work issued computer, and don't need one at home. The market for high end home hardware is really o
102.
▲
by
BobbyJo
5mo ago
> How do you run a society based on who is dissatisfied? We already do. Most of what governments do is reactionary to dissatisfaction. > then changing laws because those people are "dissatisfied" seems kind of arbitrary It&#
103.
▲
by
BobbyJo
5mo ago
Two things: 1) You have to get it out of your head that it is enough when everyone has X standard of living. It isn't. It's enough when less than a critical threshold of the population is dissatisfied, and that dissatisfaction can
104.
▲
by
BobbyJo
6mo ago
Why is China relevant?
105.
▲
by
BobbyJo
6mo ago
> In contrast, OpenAI has no such problem. It did not have CSAM pushed onto it, it actively collected such data itself. It could have, at any point before and after, simply stopped scraping all of the web indiscriminately and switched to
106.
▲
by
BobbyJo
6mo ago
The core issue with pure libertarian ideas is that they ignore "shoe leather costs" and the role they play in making consumers "irrational", which the success of a libertarian society depends on. In this case, plenty of
107.
▲
by
BobbyJo
6mo ago
I think you are inverting things. Not all anecdotes are scientific data points, but all scientific data points are anecdotes in isolation.
108.
▲
by
BobbyJo
6mo ago
Weather can be due to climate, and time series are composed of anecdotes.
109.
▲
by
BobbyJo
6mo ago
Its pretty obvious that a society intent on making capitalism work would enforce price transparency. If you want a productive society, you need efficient markets, and if you want efficient markets, you need to reduce information asymmetry,
110.
▲
by
BobbyJo
6mo ago
People just do not understand how big and spread out the US is compared to other countries. "Last mile" dramatically underestimates how much heavy lifting the personal transportation part would need to do. More like "last 50
111.
▲
by
BobbyJo
6mo ago
It's really hard to tell. Almost all the models have the benchmarks in their training data, which pushes us into the realm of basing model capability rankings on vibes. I think the OSS models tend to do worse on things outside their co
112.
▲
by
BobbyJo
6mo ago
It just makes it weird. When the US does something bad and people go "but CHINA!" it's also weird.
113.
▲
by
BobbyJo
6mo ago
... wouldn't this be a slight smidgen of evidence?
114.
▲
by
BobbyJo
6mo ago
Vietnam and Korea were technically wars to stop conquest, no?
115.
▲
by
BobbyJo
6mo ago
> Would you rather them never hire them in the first place? Isn't the obvious answer yes for everyone that sells their labor? If I gave you the choice between being an employee in an economy where it is more difficult to land a job,
116.
▲
by
BobbyJo
6mo ago
Technological advancement is speeding up. When you don't have to worry about selling your labor, it is an increasingly powerful source of comfort. When you do have to sell your labor, it is an increasingly powerful source of insecurity
117.
▲
by
BobbyJo
6mo ago
Land != Housing.
118.
▲
by
BobbyJo
6mo ago
"A significant number of jets it entered the war with" does not mean they haven't also gotten newer jets.
119.
▲
by
BobbyJo
6mo ago
I think businesses running datacenters are much less likely to frivolously buy the latest GPUs with no functional incentive than general consumers are...
120.
▲
by
BobbyJo
6mo ago
At a consistent amount of usage, datacenters are at least an order of magnitude more hardware efficient. I'm sure Nvidia and AMD would be fine fighting for B2C if it meant volume would be 10+x. Now, given they can't satisfy curren
More ›