3 ms·

The trouble right now, in the context of engineering, is that it is terribly difficult to get AI to produce something you can be confident in. We've been burned

by 9rx 2y ago

The trouble right now, in the context of engineering, is that it is terribly difficult to get AI to produce something you can be confident in. We've been burned by the human junior engineer who produces something that gets things done in the present, but two-three years down the line becomes a nightmare liability, and AI is now that same junior engineer. It is too early to tell if what AI produces will also become that terrible nightmare in a few years, or if AI will also become the saviour to make what would have been a nightmare a delight. So, one has to be in defence mode in case the nightmare reality is the one that comes to fruition, which limits how far it can be taken. Unless you just want to produce cool stuff that you plan to throw away tomorrow. That is perfectly valid! But not usually what engineers get to work on. Time will give more clarity.

skydhash 2y ago
How much time do we need? Industrial invention used to be judged by actual usefulness. If we take space exploration as an example, at every stage, we have actual usefulness for what the technology so far was able to accomplish. Same with computing. But now we're two years after the first release to the general public of LLMs and no one can provide an actual, reliable use case for it. Instead, it's mostly about destroying values, hindering workflows, and crippling the internet.
- 9rx 2y ago
  > How much time do we need? A few years will start to show if what AI is producing is a lability, or if it is just fine, methinks. If you haven't felt any pains from its legacy cruft by that point, it isn't apt to magically appear a long way down the line. > If we take space exploration as an example, at every stage, we have actual usefulness for what the technology so far was able to accomplish. When Archytas propelled a wooden bird with steam, the first recognized step towards building the rockets that got us to space, what usefulness was found in that? It offered a basic foundation on which we were able to consider other solutions and the concept in general. But couldn't you say that is, at very least, where we are at with LLMs right now too?
  - skydhash 2y ago
    Let's ignore the hype and marketing and truly consider LLM as the first stage of "something". The truth is that I've heard a lot of speculation, but no inching towards any of the speculated objective on a practical scale. For anything else, any such first stage was usually constrained on a very small scale while the glaring issues were ironed out. Now a lot of us are being forced to tolerate what is essentially a prototype or toy in tools and systems that were fine before.
    - 9rx 2y ago
      Perhaps LLMs are in the Model T stage? The experience is terrible. Its slow. The roads are muddy and full of ruts. You are bound to break your arm when you spin the crank. But, still, it offers something ever so slightly more than the horse that you put up with it anyway.
      - skydhash 2y ago
        More like a Model T that goes fast in the wrong direction, don't respond to the steering wheel, decide to reverse on random occasions,... The road may be bad and the horse is not that great, but no one would hop on that particular model unless for entertainment purposes.
        9rx 2y ago
        > that goes fast There is nothing fast about LLMs. That is their biggest pain point of all when used in an engineering context. Even if they offered perfect production, you can still do it faster yourself by a significant margin. > but no one would hop on that particular model unless for entertainment purposes. Sounds about right. It might be the "in" thing in tech circles, but the people I encounter on an everyday normal basis have never used LLMs. Many of them don't even know what an LLM is. And that's with LLMs being freely available! Imagine if you had to pay to use them. It would be a handful of people at best.
      - Seb-C 2y ago
        Natural language is a terrible interface to interact with a computer, worse than any UI or programming language. LLMs are more like replacing automobiles with mechanical horses. You fundamentally cannot get it significantly better.

by 9rx 2y ago

skydhash 2y ago

How much time do we need? Industrial invention used to be judged by actual usefulness. If we take space exploration as an example, at every stage, we have actual usefulness for what the technology so far was able to accomplish. Same with computing. But now we're two years after the first release to the general public of LLMs and no one can provide an actual, reliable use case for it. Instead, it's mostly about destroying values, hindering workflows, and crippling the internet.

9rx 2y ago
> How much time do we need? A few years will start to show if what AI is producing is a lability, or if it is just fine, methinks. If you haven't felt any pains from its legacy cruft by that point, it isn't apt to magically appear a long way down the line. > If we take space exploration as an example, at every stage, we have actual usefulness for what the technology so far was able to accomplish. When Archytas propelled a wooden bird with steam, the first recognized step towards building the rockets that got us to space, what usefulness was found in that? It offered a basic foundation on which we were able to consider other solutions and the concept in general. But couldn't you say that is, at very least, where we are at with LLMs right now too?
- skydhash 2y ago
  Let's ignore the hype and marketing and truly consider LLM as the first stage of "something". The truth is that I've heard a lot of speculation, but no inching towards any of the speculated objective on a practical scale. For anything else, any such first stage was usually constrained on a very small scale while the glaring issues were ironed out. Now a lot of us are being forced to tolerate what is essentially a prototype or toy in tools and systems that were fine before.
  - 9rx 2y ago
    Perhaps LLMs are in the Model T stage? The experience is terrible. Its slow. The roads are muddy and full of ruts. You are bound to break your arm when you spin the crank. But, still, it offers something ever so slightly more than the horse that you put up with it anyway.
    - skydhash 2y ago
      More like a Model T that goes fast in the wrong direction, don't respond to the steering wheel, decide to reverse on random occasions,... The road may be bad and the horse is not that great, but no one would hop on that particular model unless for entertainment purposes.
      - 9rx 2y ago
        > that goes fast There is nothing fast about LLMs. That is their biggest pain point of all when used in an engineering context. Even if they offered perfect production, you can still do it faster yourself by a significant margin. > but no one would hop on that particular model unless for entertainment purposes. Sounds about right. It might be the "in" thing in tech circles, but the people I encounter on an everyday normal basis have never used LLMs. Many of them don't even know what an LLM is. And that's with LLMs being freely available! Imagine if you had to pay to use them. It would be a handful of people at best.
    - Seb-C 2y ago
      Natural language is a terrible interface to interact with a computer, worse than any UI or programming language. LLMs are more like replacing automobiles with mechanical horses. You fundamentally cannot get it significantly better.