481–490 of 800 posts
(I'm not the earliest adopter of crypto and AI by any means. I only rode up crypto a couple of times for 2X and 3X kinda gains on my investment, and I only started using Claude last year.)
Feels like a false equivalency. It's just my experience, but I've completely ignored crypto and the metaverse, and I don't get the sense I'm missing out on much. In contrast, LLMs in their current state have (for me) dramatically reduced the distance between an idea and a working implementation, which has been legitimately transformative in my software dev life. Transformative for the better? Time will tell I suppose…
Sure, maybe crypto changed some lives, but an entire industry? I think ALL of software dev is going under a transformation and I think we're past the point of "wait it out" IMO.
Or I'm wrong, but right I'm being paid to develop a new skill professionally. Maybe the skill ends up not being useful - ok, back to writing code the old way then.
Earlier quoted context omitted.
What I get a bit annoyed is companies forcing AI tools, getting usage metrics and actively hunting the engineers that don't use the tool "enough", I've never seen anything like it for a technically optional tool. Even in the past, aside from technical limitaions, you were not required to use enough of a tool. It just sounds like a giant scheme to burn through tokens and give money to the AI corps, and tech directors…
As an employee of a big tech company doing this, it's all fear mongering. We are being told that if everyone doesn't use these tools, our competitors will wipe the floor with us because they are using them and will ship features 10x faster. But many engineers are suspicious as well.
Earlier quoted context omitted.
In the 90s was also the dotcom boom, and the vast majority of those who placed an all-in bet on it being everything lost it all in the dotcom bust and also "ended up being forgotten by history". Some of those bets were prescient but too early but many of those bets never made any sense. The dotcom bust was worse than the software industry crash we're experiencing now. "It's rather obvious that this AI thing is a tran…
> In the 90s was also the dotcom boom, and the vast majority of those who placed an all-in bet on it being everything lost it all in the dotcom bust and also "ended up being forgotten by history". I think you are missing the point, and also the very site you're posting on. Look at the top 50 list of most valuable companies in the world. Over half of the total market value reported today is attributed to companies whi…
But this is begging the question.
Yes, we can see that the internet was radically transformative.
But you are arguing that this somehow proves that LLMs are too, when there's wildly insufficient evidence—either on where LLMs are going in themselves, or in the comparison—to credibly make that claim.
All I know is, I've always enjoyed building things. And I enjoy building things with AI-assisted tools too, so I'll continue doing it.
Earlier quoted context omitted.
> llms are generally useful to understand more specialized / obscure topics A very simple kind of query that in my experiences causes problems to many current LLMs is: "Write {something obscure} in the Wolfram programming language."
One tendency I've noticed is that LLMs struggle with creativity. If you give them a language with extremely powerful and expressive features, they'll often fail to use them to simplify other problems the way a good programmer does. Wolfram is a language essentially designed around that. I wasn't able to replicate in my own testing though. Do you know if it also fails for "mathematica" code? There's much more text onl…
My experience concerning using "Mathematica" instead of "Wolfram" in AI tasks is similar.
There's value in being early - in the right thing. - If you'd invested in Bitcoin in 2016, you'd have made a 200x return - If you'd specialized in neural networks before the transformer paper, you'd be one of the most sought-after specialists right now - If you'd started making mobile games when the iPhone was released, you could have built the first Candy Crush Of course, you could just as well have - become an Acti…
Except you would've probably sold it at any of 1.5x, 2x, 4x, or 10x points. That's what people keep missing about this whole "early bitcoin". You couldn't tell it will 2x at 1.5x, you couldn't tell it will 4x at 2x, and so on.
“There are a 16,000 new lives being born every hour. They're all starting with a fairly blank slate. Are you genuinely saying that they'll all be left behind because they didn't learn your technology in utero?” This is a great framing.
Isn't that what is implied by the toughest job market yet for junior level candidates? Author is very confident that the answer to his question is "no".
Earlier quoted context omitted.
I'm saying that there's a cost to waiting, just like there's a cost to jumping in early. The cost is SPECIFICALLY that it is harder to jump in to a mature field with its own jargon and concerns. The assumption I think folks are making is their engineering prowess will save them here: whatever complicated thing that matters in AI land will be easily visible to an rank outsider. The whole premise of "imma wait" is not…
> The cost is SPECIFICALLY that it is harder to jump in to a mature field with its own jargon and concerns. Hasn't that been the case for decades? What specifically is different now, such that for some reason it's harder to jump in now than it was before? If anything, LLMs are supposed to make things easier, aren't they? > it's the implied "imma wait until everything falls apart, then we'll go back to what I know how…
Well, the obvious evidence that something is different now is all around us. It's been made with painful seriousness by people who thought they were making a different point, namely that LLMs represent an unreliable, poorly understood, and hazardous abstraction layer between coders and the machine. Specifically that this abstraction layer is DIFFERENT than others in the past. There are dozens and dozens of blog posts making this point (some written by machines) on HN. It would be hard to have not come across this point or miss the chorus of engineers agreeing with it. It's supposedly a cardinal reason why assembly -> C was a "good abstraction" and natural language -> slop is a "bad abstraction." If we take that argument seriously, it represents strong evidence that something new is happening, independent of anything I might say.
Why is it different? C'mon. COME ON. why can I find a post on the front page of HN when Claude is down for more than 10 minutes? Why can I find out that a new model has been released from the big 5 frontier labs, again on the front page, inside minutes? Why is it different? Did we build trillions of dollars of datacenters for NetBeans or SecondLife or whatever other cartoonish old fad I'm supposed to treat as analogous today? Are we just supposed to imagine that Microsoft, NVIDIA, Facebook, Google, Alibaba are all just staffed with idiots, or they're all caught up in irrational exuberance? Are we supposed to watch generation costs march down and outcomes improve and still think 'yeah, this is just like Pets.com? Are we supposed to yield to vague and suggestive motions toward e.g. the dot com boom as though working with agents were the same thing as investing in a specific internet company ca. 1998? Are we supposed to take from that analogy that an engineer who said "no thanks, I'll wait to see how this internet thing shakes out" in the 1990s was a real smarty to be emulated? Come on.
It's both categorically different and clearly has meaningful material force behind it.
This is a whole different interface to the computer and even if the eventual outcome is that real engineering work happens with tightly constrained and specialized harnesses around agents, understanding the actual interface is critical. Ironically, the meta-claim here is that good engineers will just be able to vibe out correct practice by engineering harder instead of understanding that core interface! Rather what will be needed is attention and orientation to concerns that people care about in the space.
I don't want to dunk on a strawman. I'd much rather not see a whole community of engineers loudly pat each other on the back for not learning about something.
Feels like a false equivalency. It's just my experience, but I've completely ignored crypto and the metaverse, and I don't get the sense I'm missing out on much. In contrast, LLMs in their current state have (for me) dramatically reduced the distance between an idea and a working implementation, which has been legitimately transformative in my software dev life. Transformative for the better? Time will tell I suppose…
> In contrast, LLMs in their current state have (for me) dramatically reduced the distance between an idea and a working implementation, which has been legitimately transformative in my software dev life. I can't really agree. I've never seen anything from an LLM that I would consider even helpful, never mind transformative. How are you supposed to use them?
I am conservative regarding AI driven coding but I still see tremendous value.
It makes me want to ask you: do you ever see helpful things from your colleagues at all?