Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
0xpgm
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
181.
▲
by
0xpgm
3y ago
What are the odds that I'm getting into chicken farming and this leads me to a chicken farm article? https://mcdim.xyz/essays/hens.html
182.
▲
by
0xpgm
3y ago
I got the sinking feeling something might have happened to Bret Victor when I saw the title. Thankfully, its just an update about his project. Maybe it can be slightly edited to reflect that.
183.
▲
by
0xpgm
3y ago
> It is an indictment of American universities that despite having the funding, resources and technology to do this a lot earlier, it was eventually a 3rd world country that pulled it off. I assume knowledge discovery and sharing doesn&#
184.
▲
by
0xpgm
3y ago
I'm just speculating, I do not 'know'. I look up to them as well. But reading about them, things that seems to stand out is that they are/were: 1) very prolific in the engineering sense. writing lots of code or producing
185.
▲
by
0xpgm
3y ago
The title is a bit click-bait, since it makes it sound like this is an official position by Google, yet the article clarifies that this the opinion of a Google employee
186.
▲
by
0xpgm
3y ago
> That's a wonderful sentiment but we said the same thing about the web and email Really? Web browsers, web servers and mail servers are beasts. I don't see them being 'simple enough to implement in a weekend, in your lang
187.
▲
by
0xpgm
3y ago
The lab pages are archived from here: https://web.archive.org/web/20141216095808/http://lynxline.c...
188.
▲
by
0xpgm
3y ago
Likewise, I thought it was something to do with blood pressure :)
189.
▲
A Retrospective on Paradigms of AI Programming (2002)
(norvig.com)
45 points
by
0xpgm
4y ago
|
33 comments
190.
▲
by
0xpgm
4y ago
> His prediction was based on the trends he was seeing at the time. But it wasn't even AI. Instead he made this prediction because he saw the rise of no-code tools replacing software developers because managers could finally cut out
191.
▲
by
0xpgm
4y ago
I starting to believe the biggest risk of AI is that it will become good enough to fool people into thinking it is sentient, and they'll start trusting it. With its inherent biases from the training data