I've always called LLMs like the Watt steam engine. LLMs are an engine designed to convert electricity into attention.
You use an engine to pump water and drill holes. Then they got more complex and moved heavy objects. Then they started powering warships. Warships were always powered by wind but the engine gave just enough power to build ships out of heavier materials and carry heavy armaments, which changed the balance of power, and led to a third of the world declaring independence from European colonial powers.
LLMs are like that. They can process a lot of text. They can think better than humans because they're doing large searches. They can be far more creative than any human. We're not there yet though, but there's a century gap between the Watt engine and the steam powered warships.
Calling it AI has had... very negative effects on how people use the technology. Instead of using it to process things, they're used on a layer that they don't belong to. People are trying to accelerate to the warship era with lots and lots of money, but money doesn't work that way.
Power tools? seems something more -high level- like: "Make me a table", not "saw this board", and evolving to: "make me a 2-level house".
Will the house be solid? You have to cheek indeed, and more complex the tool, mode depth the checking of the results, given the uncertains about the inned workings of these systems.
But correct use of these tools is IMHO the big big problem.
The -best- tool-language-tech is the one you know well, because this will avoid the use for the wrong problem, a situation very uneasy discovering ahead in development. Those AI tools are well understood? Seems not quite.
This makes me think of something that I see in the game dev space. People using AI to code and being ok with that, but strictly are against the art or music aspect of it being AI (more and more AI allows us to sit in the director's chair). People thought the same thing about Photoshop back in the day or how about auto-key framing with 3D animation instead of doing everything by hand? Or even with classic cel drawing being replaced by 3D; I remember when everyone viewed 3D animation as a cheat and not a real form of art.
Change can be very difficult, but it's here to stay whether we like it or not (with all the good and the bad that comes with it). One way or the other AI is a tool, it's unlike anything thought possible, but still a tool nonetheless. Sure, you can tell it to make generic garbage all over the place, but I would argue that a human guiding the AI working together can produce content that is truly something spectacular. This isn't to take away from how things were done before. We can/should respect the past and learn from it, but we always need to continue to move forward.
> It's just a new kind of tool for you to pick up, if you have the courage and heart to do so.
Soon it will be the only kind of tool your boss is going to pay for (if you still have a boss). And it is a tool owned by stakeholders. That sucks big time. You will never own the tool the same way you own a chainsaw. You’ll be perpetually paying for it.
A drill press doesn't spontaneously turn into a table saw that ejects its blade because although you secured the bit in the chuck you didn't secure the blade on the arbor.
AI is to software like particle board/chipboard is to old growth wood.
Particle board didn't replace all other uses of wood. They made wood products accessible to more people. They make furniture less expensive. We don't have fewer lumberjack jobs because of particle board. And with particle board came a proliferation of styles that empowered people to toss out the old furniture every few years as their tastes changed. [0]
I don't get the "agents are like factories" analogy. It sounds identical to the argument used for software development in general. That it, it takes a lot of work to produce software, but once done, essentially perfect digital copies of that one thing are effectively free.
Also, "motivated by fear" includes the fear of missing out, so the flip side of the same coin is that manipulators and influencers of the latest hype peddle it for all they are worth.
[0] leading to more waste, plus the slow release of formaldehyde.
i mean, you still guide a power tool. that requires a bit of expertise. for now, getting good products out of AI also requires expertise, but the future is unknown for that holding.
it's more about the transition from woodworking to some kind of lumber factory. a lot of people here didn't sign up to be factory managers, they wanted to work with their hands.
I've always called LLMs like the Watt steam engine. LLMs are an engine designed to convert electricity into attention.
You use an engine to pump water and drill holes. Then they got more complex and moved heavy objects. Then they started powering warships. Warships were always powered by wind but the engine gave just enough power to build ships out of heavier materials and carry heavy armaments, which changed the balance of power, and led to a third of the world declaring independence from European colonial powers.
LLMs are like that. They can process a lot of text. They can think better than humans because they're doing large searches. They can be far more creative than any human. We're not there yet though, but there's a century gap between the Watt engine and the steam powered warships.
Calling it AI has had... very negative effects on how people use the technology. Instead of using it to process things, they're used on a layer that they don't belong to. People are trying to accelerate to the warship era with lots and lots of money, but money doesn't work that way.
Power tools? seems something more -high level- like: "Make me a table", not "saw this board", and evolving to: "make me a 2-level house". Will the house be solid? You have to cheek indeed, and more complex the tool, mode depth the checking of the results, given the uncertains about the inned workings of these systems. But correct use of these tools is IMHO the big big problem. The -best- tool-language-tech is the one you know well, because this will avoid the use for the wrong problem, a situation very uneasy discovering ahead in development. Those AI tools are well understood? Seems not quite.
This makes me think of something that I see in the game dev space. People using AI to code and being ok with that, but strictly are against the art or music aspect of it being AI (more and more AI allows us to sit in the director's chair). People thought the same thing about Photoshop back in the day or how about auto-key framing with 3D animation instead of doing everything by hand? Or even with classic cel drawing being replaced by 3D; I remember when everyone viewed 3D animation as a cheat and not a real form of art.
Change can be very difficult, but it's here to stay whether we like it or not (with all the good and the bad that comes with it). One way or the other AI is a tool, it's unlike anything thought possible, but still a tool nonetheless. Sure, you can tell it to make generic garbage all over the place, but I would argue that a human guiding the AI working together can produce content that is truly something spectacular. This isn't to take away from how things were done before. We can/should respect the past and learn from it, but we always need to continue to move forward.
> It's just a new kind of tool for you to pick up, if you have the courage and heart to do so.
Soon it will be the only kind of tool your boss is going to pay for (if you still have a boss). And it is a tool owned by stakeholders. That sucks big time. You will never own the tool the same way you own a chainsaw. You’ll be perpetually paying for it.
Analogies are fun!
A drill press doesn't spontaneously turn into a table saw that ejects its blade because although you secured the bit in the chuck you didn't secure the blade on the arbor.
Analogies are fun!
AI is to software like particle board/chipboard is to old growth wood.
Particle board didn't replace all other uses of wood. They made wood products accessible to more people. They make furniture less expensive. We don't have fewer lumberjack jobs because of particle board. And with particle board came a proliferation of styles that empowered people to toss out the old furniture every few years as their tastes changed. [0]
I don't get the "agents are like factories" analogy. It sounds identical to the argument used for software development in general. That it, it takes a lot of work to produce software, but once done, essentially perfect digital copies of that one thing are effectively free.
Also, "motivated by fear" includes the fear of missing out, so the flip side of the same coin is that manipulators and influencers of the latest hype peddle it for all they are worth.
[0] leading to more waste, plus the slow release of formaldehyde.
i mean, you still guide a power tool. that requires a bit of expertise. for now, getting good products out of AI also requires expertise, but the future is unknown for that holding.
it's more about the transition from woodworking to some kind of lumber factory. a lot of people here didn't sign up to be factory managers, they wanted to work with their hands.