Comment by onlyrealcuzzo
15 days ago
Assuming that models continue to get 90% more efficient every 18 months, and assuming that hardware continues to get 2x more powerful per watt every 2 years...
Aren't we still like a decade away from an Argon model being able to run on battery power for any reasonable amount of time that would fit on a robot that isn't the size of a school bus?
Add on another couple of years for the speed to get to what it likely needs to be to be generally useful instead of narrowly useful?
I mean - I guess we're maybe 5-7 years away from a reasonably priced robot you could plug into the wall like an old vacuum cleaner (that runs on this type of tech).
I think for 95%+ of applications.. you just need a link to a data center anywhere.
My assumption is the hardware will have one/several smaller models running on it for locomotion, balance, etc.
Then the hardware "phones home" for decision making and planning.
Essentially the planning doesn't need to be close to the device as long as the local models can handle some part of the plan execution and make sure the robot doesn't fall over. The latency shouldn't be an issue. In fact more powerful centralized compute quickly creates an advantage over slower local compute in terms of end to end planning.
I can see some liability issues with that approach.
Liability issues will exist regardless of whether compute is on-device or central.
However, if the robot is in a contained environment like a home, there is less third party liability than for something like vehicles. Not none, of course, but less. And when you can get your arms around third party liability, first party liability can be better controlled by contractual agreements.
In general I expect on-device motor execution will allow it to safely come to a stop or enter a safe position if it loses the link to the execution planning cloud model.
1 reply →
> Aren't we still like a decade away from an Argon model being able to run on battery power for any reasonable amount of time that would fit on a robot that isn't the size of a school bus?
Sooner or later, the models will stop improving. After that happens, the best models will be etched into model-specific ASICs with weights fixed in mask ROM (like Taalas is already doing), which will result in massive, at least order of magnitude inference efficiency increase.
They said the same thing for transistor density in new chip nodes. It took 40 years of doubling every 2 years to finally hit a wall.
And although the industry hit a wall in density/area, they switched to scaling in 3D or glueing chiplets together. Moore's Law is still working if you change the definition to transistors/integrated circuit.
Transistors doubled for so long, but pretty much no other technology has come close, so it’s not like going to improve for very long, which is far more common. Transistors only did it due to a quirk in physics, that of electrons being long lasting, super tiny, and fairly controllable. We’re out of such particles to make another set of leaps.
For data bound things, there’s also fundamental laws of information theory, far stronger than physics theories, since they’re theorems about math. And LLMs can only compress knowledge so far, and we’re about at it now.
That doesn’t help the RAM situation
I'm naive, but wouldn't industrial applications be enabled by this anyway? You'll have in-house inference hardware that a robot can talk to via local network.
Also, your model probably doesn't need like a lot of its knowledge to control a robot.
That's only bringing up the horizon maybe 1 year. At industrial scale - I guess something like an Amazon warehouse could start to be interesting with a local datacenter IN the warehouse within 2-3 years.
From a cost perspective, I just can't imagine that heavily, heavily specialized models wouldn't crush a general model, though - which Amazon already does.
Everyone thinks specialized models are the ticket, except then they aren't: http://www.incompleteideas.net/IncIdeas/BitterLesson.html
3 replies →
You can already run top of the line models on something that isn't the size of a school bus. The couple of KW you need to run them are the bigger problem, though on very high end hardware you might get double the efficiency.
Robots stand-alone AI do not really benefit from things like batch mode so you want to keep your LLM as high level as possible and your robot's subsystems as autonomous as possible, very light coupling between the two, definitely nothing time-critical. And that's the challenge: in robotics almost everything is time critical.
Probably not many applications require the full AI compute to be local on device (Self-driving cars and military applications come to mind).
And? 5-7 years is nothing at the scale of our modern society and the outcome of what you describe could be enormous.
I don't disagree.
It just seems like jumping the shark a little bit.
It's basically completely unviable for a few years still.
It's viable today via network.
I don't think it's viable in the sense that they are still in a demo stage.
Also, I would assume these models are going to get bigger before they get smaller (but they will get smaller over the long run)
1 reply →