The cynical-but-likely-accurate take is that researching language modeling has a higher ROI and lower risk than researching robotics.
My guess is they see they are close to something very big with language models and want to invest everything there
OpenAI disbands its robotics research team
91–100 of 128 posts
Re: OpenAI disbands its robotics research team
#92Re: OpenAI disbands its robotics research team
#93Earlier quoted context omitted.
Even if this is true, that weakens your point that hardware is the fundamental limit for robots. If there are situations where software giving human like behavior to a robot could be extremely valuable, then there's certainly a motivation for generic AI companies to be in that area. That doesn't mean OpenAI robotics leaving wasn't a good idea. It seems like it was but for other reasons.
Giving robots human-like behavior is mostly useful for general-purpose robots. Specialized robots don't need general AI to do their jobs. OpenAI is trying to develop general human-like AI. There's no general-purpose robot for them to put it in, and developing one is a hardware problem, not a software problem.
We can build perfectly good robot arms for a huge swath of assembly/warehouse/retail tasks, but there's no AI that can aim them well enough and carefully enough. An overqualified AI would still be a valid solution and extremely valuable.
Re: OpenAI disbands its robotics research team
#94Earlier quoted context omitted.
Even if this is true, that weakens your point that hardware is the fundamental limit for robots. If there are situations where software giving human like behavior to a robot could be extremely valuable, then there's certainly a motivation for generic AI companies to be in that area. That doesn't mean OpenAI robotics leaving wasn't a good idea. It seems like it was but for other reasons.
Giving robots human-like behavior is mostly useful for general-purpose robots. Specialized robots don't need general AI to do their jobs. OpenAI is trying to develop general human-like AI. There's no general-purpose robot for them to put it in, and developing one is a hardware problem, not a software problem.
Specialized robots don't need general AI to do their jobs.
Self-driving cars have been unable to succeed based on their lack of a broad understanding of what's happening on the road (ie, "too many corner cases"). Self-driving cars would be a huge change and their failure is very significant.
Re: OpenAI disbands its robotics research team
#95Earlier quoted context omitted.
In a nutshell it’s too wasteful in energy spent and it doesn’t even try to mimic natural cognition. As physicists say about theories hopelessly detached from reality - “it’s not even wrong”. The achievements of RL are so dramatically oversold that it can probably be called the new snake oil.
I'm going to need you to unpack that a bit. Isn't interacting with an environment and observing the result exactly what natural cognition does? What area of machine learning do you feel is closer to how natural cognition works?
Adding to the other comment, it's quite clear that animals, and especially humans, act and learn based on many orders of magnitude less experiences than pure RL needs, especially when discussing higher order behaviors. We obviously have some systems that use inductive and deductive reasoning, heuristics, simplistic physical intuitions, agent modeling and other such mechanisms, that do not resemble ML at all.
I would say that it is likely, intuitively, that these systems were trained through things that look much like RL in the millions of years of evolution. But that process is obviously not getting repeated in each individual organism, who is born largely pre-trained.
And for any doubt, the poverty of the stimulus argument should put it to rest, especially when looking at simpler organisms than vertebrates, which can go from egg to functional sensing, moving, eating, predator avoiding in a matter of minutes or hours.
Re: OpenAI disbands its robotics research team
#96I think most people believe that the problem with robots is that we don't have the right software, and if we just knew how to program them then today's robots could be incredibly useful in everyday life. From that perspective, this move from OpenAI seems dumb. That belief is wrong. Today's robots can't be made useful in everyday life no matter how advanced the software. The hardware is too inflexible, too unreliable,…
I'm just imagining a sentient AGI rolling around in the human equivalent of a toaster, trying to bend the world to its will.
Re: OpenAI disbands its robotics research team
#97Earlier quoted context omitted.
Yes, it feels like we have squeezed most of the performance out of current algorithms and architectures. OpenAI and deepmind have thrown tremendous compute against the problem with little overall progress (overall, alpha go is special). There was a big improvement in performance by bringing in function approximators in the form of deep networks. Which as you said can scale upwards nicely with more data and compute. I…
> alpha go is special The VC community is in denial about how much Go resembled a problem purpose built to be solved by deep neural networks.
Are you suggesting that Go literally was purpose built for this?
Re: OpenAI disbands its robotics research team
#98Earlier quoted context omitted.
I'm going to need you to unpack that a bit. Isn't interacting with an environment and observing the result exactly what natural cognition does? What area of machine learning do you feel is closer to how natural cognition works?
> What area of machine learning do you feel is closer to how natural cognition works? None. The prevalent ideas in ML are a) "training" a model via supervised learning b) optimizing model parameters via function minimization/backpropagation/delta rule. There is no evidence for trial & error iterative optimization in natural cognition. If you'd try to map it to cognition research the closest thing would be behaviorist…
just to say this is the kind of answer that makes HN an oasis on the internet.
Re: OpenAI disbands its robotics research team
#99I think most people believe that the problem with robots is that we don't have the right software, and if we just knew how to program them then today's robots could be incredibly useful in everyday life. From that perspective, this move from OpenAI seems dumb. That belief is wrong. Today's robots can't be made useful in everyday life no matter how advanced the software. The hardware is too inflexible, too unreliable,…
I don't think it's really true to say that the issue is hardware or software. There is lots of robotics in everyday life, from autonomous vacuums in our homes to autonomous factories producing the goods we consume. The reason we don't have millions of little robots buzzing around us is... there's very little need for it. The average human spends most of their time barely engaged, our brains and bodies are operating f…
Can you be more specific about what the social problem is? Do you know of anyone who is working to solve it?
Re: OpenAI disbands its robotics research team
#100This is lunacy. The first country/company to replace human labour with general bipedal robots, will reap wealth beyond imagination. The short sitedness is astonishing, if you ask me. I genuinely believe how we as a society act once human labour is replaced is first aspect of the great filter.
You might be right long term - but think about it short term. Labor in low cost countries is very cheap. A few thousand a year. Unless you can build these machines for low 10's of thousands and maintain them for 100's per year, the economics won't work. Construction robots might be a good counter example because you can't offshore them.