jmathai

16 ms·

301.
by jmathai 2y ago
It felt like public infrastructure. I don’t think anyone mistook it as such but its openness probably made many of us forget, at times, that it was private and things could change in an instant.
302.
by jmathai 2y ago
This same approach can work for eliminating errors in LLM generated code. We've had good luck doing it with Lattice [1] going from 5% to 90% success rate of creating a fully functional application. I'm drafting a blog post that ta
303.
by jmathai 2y ago
I was very surprised the first time I saw JSX when working on React code. It really felt like a bastardization of what my gut felt were good software engineering practices. But I concluded that surely these devs and Facebook are smarter tha
304.
by jmathai 2y ago
This is something I call latent product development. Many times, the best solution is not the first one you think of - even though it may be one of many correct solutions. I have found that the best products and product features have come f
305.
by jmathai 2y ago
Banner ads back then were more obnoxious. Today they are deceptive. The point being made here is that it wasn’t perfect before but for many it was better.
306.
by jmathai 2y ago
Oh man. Pop under ads. Those sucked. I’d still take the old Internet with pop under ads over what we have today.
307.
by jmathai 2y ago
Some people making money on the Internet is a lot different than what the Internet has become today - and what I meant by money grab.
308.
by jmathai 2y ago
Most people don't remember, and some have never experienced, the Internet before it became a money grab. I think a lot of people outside of HN would prefer that Internet way more than what we have now.
309.
by jmathai 2y ago
This is an interesting use case for LLMs. At first, I thought “why complete an MVP if you don’t have time to iterate it?”. But MVPs are a good testbed for determining if something is worth iterating. I build quick prototypes for validation
310.
by jmathai 2y ago
Also building something in this space. I think it’s a mistake to compare the speed of LLMs to humans. People don’t like to sit and wait. The more context you can give the better but at some point (>30 seconds) people grow tired of waitin
311.
by jmathai 2y ago
We use the approaching of feeding mistakes from LLM generated code back to the LLM until it produces working code [1]. I might have to try some more aggressive prompting :). [1] https://withlattice.com
312.
by jmathai 2y ago
From the article, there seemed to be evidence of mental instability and propensity towards violence. I don’t know the line but it seems this would obviously warrant investigation regardless of if this person broke a law. I say that as someo
313.
by jmathai 2y ago
Correct. Making core Firefox better is not going to generate revenue. Not sure what people expect, every bold new idea (I.e. mobile OS) is chasing something. And they are trying to do it while being open. You’d think this HN crowd has never
314.
by jmathai 2y ago
What site is this? I miss some sites from 15+ years ago which let people post bootlegs.
315.
by jmathai 2y ago
Seems reasonable to implement a portion of an open spec for niche use cases like this. Every browser takes several gigs of ram once you have more than a couple tabs open. I’m in the niche looking to trade off some functionality to gain some
316.
by jmathai 2y ago
I don’t think they are in the majority, but it doesn’t have to be typical social activities like sports. I spend time WFH dropping and picking up my 3yo from pre school. In doing so I meet a number of parents from the community.
317.
by jmathai 2y ago
https://archive.is/2024.12.21-093402/https://www.wsj.com/tec...
318.
by jmathai 2y ago
https://archive.is/2024.12.21-093402/https://www.wsj.com/tec... I read a lot of these posts and get the impression that they are trying to make the point that AI isn’t transformational. The current suite
319.
by jmathai 2y ago
I may be in the minority, but I don’t see any appeal to AI generated art. I’m not sure which positive emotion it is supposed to evoke. I also dislike heavily edited photos for the same reason. At some point it is no longer real. That point
320.
by jmathai 2y ago
I agree with all of that. Perhaps what I meant to say was that these weren't the things users were complaining about. I love a beautiful UI and experience. Listen to users and prioritize what helps them.
321.
by jmathai 2y ago
It's not always the right tool but in more cases than modern devs believe, it is . I've built several web applications recently that don't use any modern JavaScript framework. You'd be surprised how quickly a page can
322.
by jmathai 2y ago
I understand there's a lot of hype around "build apps quickly without coding". I have many doubts as to the validity of the Youtube speed runs which depict someone actually doing it. Cursor is a developer productivity tool -
323.
by jmathai 2y ago
> Any particular reason you stopped working on it? I wasn't seeing a viable path to meaningful revenue. It looks like you may have (or are approaching) cracking that - awesome!! > but I like the way you show deltas in table cells
324.
by jmathai 2y ago
This looks great - congrats on launching it and getting it to where it is today. I launched a personal financial forecast tool a few years ago with a primary focus being to help people understand the draw down plan. This is not an ad for my
325.
by jmathai 2y ago
This is a very good starting list.
326.
by jmathai 2y ago
Clean code is like #5 or lower on the list.
327.
by jmathai 2y ago
It's a test in the sense that it's meant to validate functionality. You're correct there. The endpoints we poke at are provided as context when creating the application. Our approach evolved to be more liberal in what was r
328.
by jmathai 2y ago
What is working better for us is to spin the application up and poke at it with HTTP requests and look for non 200 response codes. The tradeoff is it's not as comprehensive. But giving users something that is 95% working and making it
329.
by jmathai 2y ago
I've been working on a prompt to application product[1] and one of the approaches we tried was test driven development. We would have the LLM write tests based on a detailed description of the application. Then give the LLM the tests a
330.
by jmathai 2y ago
Never occurred to me that the response changes based on load. I’ve definitely noticed it seems smarter at times. Makes evaluating results nearly impossible.

More ›