
Natural Language Robot by 2030?
Plus
20
แน5362030
82%
chance
1D
1W
1M
ALL
Consumer price point
Get แน1,000 play money
Sort by:
As in, available for consumer purchase or at consumer price points? What are consumer price points, if so?
Like, I can imagine someone making a basic robot (like the roomba mentioned by Jon Simon) that is pretty cheap and then just hooking up to a LLM with the outputs of the LLM hooked up to cleaning-area prioritization so you can tell it to prioritize cleaning the spill in front of the TV easily. But it could not be official, and would be aftermarket modification with opensource mod.
This is clickbait. Parameters does not mean intelligence. We don't actually know whether PaLM is good or not, but my bet would be that it's worse for most general tasks than GPT-4.
Related questions
Related questions
Natural Language Robot by 2030?
79% chance
Will 500K humanoid robots be manufactured before 2031?
73% chance
Will there be realistic AI generated video from natural language descriptions by the start of 2026?
46% chance
Will the 3 laws of robotics be used as a framework for an AGI or advanced AI by 2030?
20% chance
Will AI translators be as good as humans before 2030?
89% chance
By the start of 2026, will I still think that transformers are the main architecture for tasks related to natural language processing?
68% chance
What smart humanoid bots' capabilities will be showcased in videos by 2030?
Apple Releases Humanoid Robot by 2040
35% chance
Will 100 million humanoid robots have been produced by 2035?
33% chance
Will 200K humanoid robots be manufactured before 2030?
68% chance