Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
padolsey
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
21 ms
·
301.
▲
by
padolsey
3y ago
I've had more luck with getting it to output XML as (1) You can imbue XML with actual language/meaning (which LLMs adore) and (2) parsers can be made to be more forgiving. I get why people want to make JSON, but to me it's a
302.
▲
by
padolsey
3y ago
Ah that's fair. But it is my impression that the bulk of privacy/confidentiality concerns (e.g. law/health/..) would require "end to end" data safety. Not sure if I'm making sense. I guess microsoft is som
303.
▲
by
padolsey
3y ago
I'm confused. If this is just a front-end for the OpenAI API then how does it remove the data privacy concern? Your data still ends up with Azure/OpenAI, right? It doesn't stay localized to your instance; it's not your G
304.
▲
by
padolsey
3y ago
It seems to reflect the general way we understand the brain right? Wiring together/firing together? Then ~ abra cadabra ~ meaningful blobs of brain buzzy stuff emerge from seemingly simple rules? It seems beautifully pure that mind map
305.
▲
by
padolsey
3y ago
Love it!! The chunking stuff especially is really impressive. Hitting those token limits often is the annoying bit of working with LLMs. A weird question: How do you feel about possibly ~wasted efforts of these techniques when gpt in a year
306.
▲
by
padolsey
3y ago
Very much agree with what you've said here, and I love the idea that it makes me a better programmer/mentor. Hadn't really thought of it that way!
307.
▲
by
padolsey
3y ago
Hmm fair. It's strange that our experiences are so different. Can I ask what types of problems you ask it to solve? FWIW I've had to take quite a lot of time figuring out how to talk to it in a way that gets good results.
308.
▲
by
padolsey
3y ago
> Instead, you head over to Reddit where the programming community is much nicer Is this true? Wondering if there's a more objective way to know than endless anecdotes. I think programming communities stereotypically are pretty mea
309.
▲
by
padolsey
3y ago
Could this time be different? The tools are now in the hands of the "masses", not behind closed doors or in lofty ivory towers. People can run this stuff on their laptops etc
310.
▲
by
padolsey
3y ago
Oh I know I'll get over it, as with all things. Still in that grieving place tho :)
311.
▲
by
padolsey
3y ago
Would be nice if you could further qualify that statement.
312.
▲
by
padolsey
3y ago
For myself at least, Twitter has been instrumental in my career, hobbies and friendships. But now, seeing it flounder, and having people shard off to mastadon, bsky, and siloed instances of [insert next federative platform], makes me so ver
313.
▲
by
padolsey
3y ago
Hmmm, I don’t think that's accurate. To label plant ecosystems as 'winner-takes-all' really misses the forest for the trees (sorry..!). It's not all about sun-hogging; different species have unique needs and abilities, t
314.
▲
by
padolsey
3y ago
> what libs / functions to use, how to call them Yeh same. It's got pretty good overview of what libraries are available. I tend to ask it for an npm module to do x and it always has a couple options, and can list pros/con
315.
▲
by
padolsey
3y ago
> the code can almost write itself My 2c: I've been an eng for around 15 yrs. I semi-recently had a brain injury so haven't been able to dedicate anywhere near as much mental cognition to programming recently. That's why I
316.
▲
by
padolsey
3y ago
Dude I’m literally—right now— using an LLM to fuel and *build* new features in my production website (ablf.io). I’m just some random eng. I’m sure there are many like me. And I’m not talking about gpt just recommending the odd function. It
317.
▲
by
padolsey
3y ago
I don’t understand why this talking point keeps surfacing. LLMs are in significant usage already in many production environments and are already improving peoples’ lives and productivity. I can have a dig for more examples but I’d urge you
318.
▲
by
padolsey
3y ago
Ambition is a pretty ambiguous term for what I think—here—means “A strong yearning for a type of success characterised by western capitalist-individualistic schema of wealth and status.” Cool if you want that I guess. But it’s narrow af.
319.
▲
by
padolsey
3y ago
Biased,.. but I've been building a site that does this for a few years now and reckon it's not bad. It's more of a passion project. Lets you type in books you like and then will find the most intersecting set of other books.
320.
▲
by
padolsey
3y ago
What upsets me most is how much of a treasure trove of data that amazon/goodreads are sitting on and yet their recommendations are still pitiful. A basic collaborative-filtering algorithm would be better than "OH WE THINK YOU WILL
321.
▲
by
padolsey
3y ago
AFAIK his paintings were never intended to be 'realistic' in the sense we may value, but instead: symbolic. He played with color, intensity, and perspective, expressing vast emotion, and in some cases, turmoil. His mental health a
322.
▲
by
padolsey
3y ago
> I haven't seen or heard of anyone that seriously uses ChatGPT to generate code and uses it in prod I have. And I'm sure others have too. I dare say there are many who probably _shouldn't_ be using it because of privacy&#
323.
▲
by
padolsey
3y ago
> The TLDR here is that PMs should not be running all project meetings. Engineering managers, IC project leads, or tech leads need to be managing engineering execution, not PMs. I think this is a big one. Actually, in a way, maybe PMs sh
324.
▲
by
padolsey
3y ago
A startup still needs capital, even in the era of increasingly cheap LLMs. And the bar of innovation is now higher. The playing field has changed; if everyone now has access to increasingly competent AIs, that means the value–and differenti
325.
▲
by
padolsey
3y ago
Once the temporal fixes are applied, the residual morphing and flickering actually quite appeals to me. Reminds me of the animation style in A Scanner Darkly.
326.
▲
by
padolsey
3y ago
Looks really good - pleasant UX and seems nicely semantic/accessible! Also seems to allow slight mispellings which I really like. I tried building my own autocomplete thing for ablf.io and it was *tough*. The dataset was massive, over
327.
▲
by
padolsey
3y ago
I think, for me, it is not a product. It is a creation. There inlies the crux i suppose...
328.
▲
by
padolsey
3y ago
I wrote a book, and remember reading the first review. It was a one-star rating complaining that the book had been published late. I was in hospital at the time recovering from a brain injury. My brother had helped get the book over the fin
329.
▲
by
padolsey
3y ago
I have always taken feedback poorly, and my brain has been very unkind to me during performance reviews. I don't really trust people who claim to "love feedback" or find it "good for personal growth". It sounds so f
330.
▲
by
padolsey
3y ago
It's very cool, and uses the same principles as lots of other neurorehab electrical devices out there. I used an FES ('Functional Electrical Stimulation') device for a while that has a little accelerometer in it to figure out
More ›