Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What could have been

coppolaemilio.com

Post reply on HN
Print view View original

61–70 of 127 posts

Re: What could have been

#61

Post by pixl97 »

>There isn’t a single day where I don’t have to deal with software that’s broken but no one cares to fix

Since when does this have anything to do with AI? Commercial/enterprise software has always been this way. If it's not going to cost the company in some measurable way issues can get ignored for years. This kind of stuff was occurring before the internet exists. It boomed with the massive growth of personal computers. It continues to today.

GenAI has almost nothing to do with it.

Re: What could have been

#62

Post by simonw »

roxolotl wrote: post #49
simonw wrote: post #21

Earlier quoted context omitted.

LLMs offer the single biggest advance in interoperability I've ever seen. We don't need to figure out the one true perfect design for standardized APIs for a given domain any more. Instead, we need to build APIs with just enough documentation (and/or one or two illustrative examples) that an LLM can help spit out the glue code needed to hook them together.

The problem with LLMs as interoperability is they only work sub 100% of the time. Yes they help but the point of the article is what if we spent 100billion on APIs? We absolutely could build something way more interoperable and that’s 100% accurate. I think about code generation in this space a lot because I’ve been writing Gleam. The LSP code actions are incredible. There’s no “oh sorry I meant to do it the other wa…

If you gave me $100 billion to spend on API interoperability, knowing what I know today, I would spend that money inventing LLMs.

Re: What could have been

#63

Post by jibal »

busterarm wrote: post #40
MyOutfitIsVague wrote: post #38

Earlier quoted context omitted.

Been here for years (across many different accounts), and this is the first time I've heard of these terms. I am here for programming content, not business.

Why should I expect anyone who openly admits to having multiple accounts on the site in clear opposition to the site rules to have any level of awareness? I'm sure dang will come and ding me for this one, but I'm sitting here having my points undermined by literal sockpuppets.

Weird ad hominem flex. And "across" doesn't imply multiple accounts at once. People who register with their company email may be forced to create a new account when they leave the company (perhaps with sudden and unexpected loss of access to the account).

Here's the actual guideline (not rule):

"Throwaway accounts are ok for sensitive information, but please don't create accounts routinely. HN is a community—users should have an identity that others can relate to."

People have commented both appreciating your clear definitions and calling you out for the condescension, with a perfect xkcd suggesting an attitude change. It's up to you how you react to such feedback.

Re: What could have been

#64

Post by dougdonohoe »

pixl97 wrote: post #61

>There isn’t a single day where I don’t have to deal with software that’s broken but no one cares to fix Since when does this have anything to do with AI? Commercial/enterprise software has always been this way. If it's not going to cost the company in some measurable way issues can get ignored for years. This kind of stuff was occurring before the internet exists. It boomed with the massive growth of personal comput…

The point is that money that is going into GenAI or adding GenAI-related features to software should be going to fix existing broken software.

Re: What could have been

#65

Post by hoytie »

pixl97 wrote: post #61

>There isn’t a single day where I don’t have to deal with software that’s broken but no one cares to fix Since when does this have anything to do with AI? Commercial/enterprise software has always been this way. If it's not going to cost the company in some measurable way issues can get ignored for years. This kind of stuff was occurring before the internet exists. It boomed with the massive growth of personal comput…

I think the point the author is trying to make is that there are many problems in plain sight we could be spending our efforts on, and instead we are chasing illusory profits by putting so many resources into developing AI features. AI is not the source of the issues, but rather a distraction of great magnitude.

Re: What could have been

#66

Post by gorpy7 »

epistasis wrote: post #8

While I'm somewhat sympathetic to this view, there's another angle here too. The largesse of investment on a vague idea means that lots of other ideas get funding, incidentally. Every VC pitch is about some ground-breaking tech or unassailable moat that will be built around a massive SAM; in reality early traction is all about solving that annoying and stupid problem your customers hate doing but that you can do for…

it’s peculiar because i love to use chat gpt to fill my knowledge gaps as i work through solutions to building and energy problems that i want to solve. i wonder how many people are doing something similar and, although i haven’t* read through all the comments, i doubt much is being said let alone giving credence to that simple but potentially profound idea. learning amplified.

Re: What could have been

#67

Post by Yhippa »

emp17344 wrote: post #22
ninetyninenine wrote: post #4

I disagree. The hype is wearing people down and making it think it's a waste of time but LLMs just came out a couple years back and even the trendline from the past decade (pre-LLMs) is up up up. The amount of interest to explore this opportunity is worth it. The bubble is worth it. I don't think it's lost years, and even if it is, the technology is compelling enough to make the gamble worth it. The fatigue of readin…

Of course fairly quick progress was made - a truly astounding amount of money was poured into this industry in a short timeframe. The thing is, now it’s clear that AI isn’t really valuable enough to justify investment on the same scale anymore.

It feels like after people were still flush with cash at the end of the pandemic, reality hit and as people were profit taking from the market, LLMs seemed to emerge from the æther as the next best thing to glom on to. So now the hive mind dumped all their money into that and we are riding an incredible bubble.

So cheap gaming hardware in the future (similar to when telecoms over invested in transcontinental undersea fiber-optic cables)? What's the hangover gonna look like after this? What's the next grift?

Re: What could have been

#68

Post by Uehreka »

This is another genre of AI article that annoys me: The one where the author starts by agreeing with you that AI is a definitely a bubble, and we’re gonna just “know” that for the rest of the article, no argument necessary.

I don’t feel like this article is trying to start a conversation, it wants to end the conversation so we can have dessert (aka, catastrophizing about the outcome of the thing “we know” is bad).

Re: What could have been

#69

Post by Yhippa »

jay_kyburz wrote: post #13
ninetyninenine wrote: post #4

I disagree. The hype is wearing people down and making it think it's a waste of time but LLMs just came out a couple years back and even the trendline from the past decade (pre-LLMs) is up up up. The amount of interest to explore this opportunity is worth it. The bubble is worth it. I don't think it's lost years, and even if it is, the technology is compelling enough to make the gamble worth it. The fatigue of readin…

I'm a massive AI skeptic, and I think the amount of money being spent is astonishing, but I really don't want to go back to searching the web the old way. Asking Gemini _is_ just much better at finding you the answers you need, _and_ providing links for you to verify that information. It will be a sad day when they start injecting ads, I really hope the foss alternatives catch up.

I still don't trust the non-determinism of current LLMs. I feel like I can't trust the results unless they are very simple ones.

Re: What could have been

#70

Post by socalgal2 »

pixl97 wrote: post #61

>There isn’t a single day where I don’t have to deal with software that’s broken but no one cares to fix Since when does this have anything to do with AI? Commercial/enterprise software has always been this way. If it's not going to cost the company in some measurable way issues can get ignored for years. This kind of stuff was occurring before the internet exists. It boomed with the massive growth of personal comput…

> Commercial/enterprise software has always been this way

All software is this way. The only way something gets fixed is if someone decides it's a priority to fix it over all the other things they could be doing. Plenty of open source project have tons of issues. In both commercial and open source software they don't get fixed because the stack of things to do is larger than the amount of time there is to do them.

Post reply on HN