Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
z5h
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
61.
▲
by
z5h
3y ago
I’ve had to deal with a nightmare Elm project. The difference is, I know what I’m doing. I solved all the problems, created examples to lead other devs and everything was fine. Have your company contact me if you need help.
62.
▲
by
z5h
3y ago
No. (Shame). But I did just download “UNIX: A History and a Memoir”.
63.
▲
by
z5h
3y ago
Great read. This made me curious about “the story” of the creation of the internet. ChatGPT suggested some books: 1. Where Wizards Stay Up Late: The Origins Of The Internet 2. The Innovators: How a Group of Hackers, Geniuses, and Ge
64.
▲
by
z5h
3y ago
I REALLY REALLY like this. This is gold. I'd buy a polished-up version of this on iOS.
65.
▲
by
z5h
3y ago
If it’s this easy to destroy decades of work, it’s not the janitor’s fault.
66.
▲
by
z5h
3y ago
Point being, why consider the outcome of GPT’s ability to dress up an idea to be persuasive, without considering its ability to cut through persuasive language and consider the core idea?
67.
▲
by
z5h
3y ago
“ Via a survey of 250 investors and 250 business owners…” Who cares what humans think. What did GPT-4 think about the pitches?
68.
▲
by
z5h
3y ago
Agreed. The original Animoog iOS synth came out in 2011, and was (still is) one of the best sounding and most innovative iOS synths. (Of course they have a V2 and many other apps out there now.)
69.
▲
by
z5h
3y ago
I’m suggesting we (AI) can find alternative and preferable “games”.
70.
▲
by
z5h
3y ago
I’m suggesting you’re not a position to make claims about a system, that in its infancy, can outperform you (most of us) in many ways. Especially if, in your opening argument, you admit being unable to think of scenarios instead of doing th
71.
▲
by
z5h
3y ago
As a species, we need to commit to the belief that a powerful enough AI can prevent/avoid/vanquish any and all zero-sum games between any two entities. Otherwise we commit to adversarial relationships, and plan to use and develop
72.
▲
by
z5h
3y ago
ChatGPT can give plausible problematic scenarios where you (admittedly) cannot. Reflect on that.
73.
▲
by
z5h
3y ago
I was looking for the modular synth comment(s). That's immediately what came to mind when I saw this. How can I integrate this with my synths.
74.
▲
by
z5h
3y ago
Is Prolog so low as to not appear?
75.
▲
by
z5h
3y ago
What is it?
76.
▲
by
z5h
3y ago
You can embed a microKanren in whatever language you like.
77.
▲
by
z5h
3y ago
Were you able to make use of the fixpoint optimization ideas in Datafun? I thought it was a great idea but the project seems to have stalled.
78.
▲
by
z5h
3y ago
I'm just paying a lawyer out of pocket that came recommended through a trusted contact.
79.
▲
by
z5h
3y ago
Yes BUT… suppose you file a provisional patent, and do a slight pivot/change in how you do things. Then the provisional patent might not apply, or might be too restrictive. This is where a lawyer can help. Still, this is a strategy tha
80.
▲
by
z5h
3y ago
As a solo creator, I've been working with a lawyer to create some intellectual property. I'm not a lawyer. This is not legal advice. Here are some options as I understand them, but do your own research. A utility patent covers a n
81.
▲
by
z5h
4y ago
I’ve been using Prolog a bunch recently, and also embedded and extended MicroKanren in a project. Something I came to appreciate was that Prolog’s depth-first search, and Kanren’s lazy stream approach are good with memory even when generati
82.
▲
by
z5h
4y ago
https://open.spotify.com/episode/1rOLlmSIlQ44MrtQ0PaSO9?si=7...
83.
▲
by
z5h
4y ago
As someone who has done commercial and personal work in Elm for many years, I know that lots of people are actively and productively using it with fewer issues than most other mainstream languages. Of course there are issues in any language
84.
▲
by
z5h
4y ago
Read “Denial of Death”. Ideas here are based on that book. The plain reality of existence is that there is no inherent meaning and we will all die. That is too much for anyone to bear. We can live “normal” lives only so far as we can distra
85.
▲
by
z5h
4y ago
It recently generated a response with incorrect English grammar. I asked if it could find the grammatical error in its response. It did, and corrected it.
86.
▲
by
z5h
4y ago
As others have suggested: seek professional help. But also, the thing I try to recall when I feel the most "stuck", is that despite how anything external to me plays out, I will always have the opportunity for inner growth. I alwa
87.
▲
by
z5h
4y ago
I gave an Elm talk years ago at the shared office space I used to work at. A few devs there later told me the talk helped them better understand React.
88.
▲
by
z5h
4y ago
> Most people desire to be happy. Right. Of course. Evolution has designed us to desire to be happy, but to be unhappy (to some degree, at some times). Otherwise we'd have no drive to do anything ever. Some degree of unhappiness
89.
▲
by
z5h
4y ago
We haven't established that intelligence should imply happiness. We haven't even established that intelligence should imply the desire to be happy. If intelligence enables a person to identify problems, and problems make a person
90.
▲
by
z5h
4y ago
I switched to this also year(s) ago and am happy with this system. I have no urge to switch.
More ›