2 ms·

I’m a skeptic. There is no evidence behind this tweet that the technique works. I may be mistaken, but I don’t see the authors as having a background in the fie

by lumost 4y ago

I’m a skeptic. There is no evidence behind this tweet that the technique works. I may be mistaken, but I don’t see the authors as having a background in the field. Likewise they don’t describe the technique in any detail. Sounds like snake oil to me, or fake it until you make it.

singularity2001 4y ago
I tried something similar with terminal copilot yesterday. The approach works for bash calls, why not for web calls?
famouswaffles 4y ago
GPT3 has demonstrably been able to run/execute execute and make API requests with python for quite some time now. You can test it too. See here https://twitter.com/sergeykarayev/status/1569377881440276481 https://twitter.com/sergeykarayev/status/1569377881440276481 https://twitter.com/sergeykarayev/status/1570868002954055682 https://twitter.com/sergeykarayev/status/1570868002954055682 None of this seems like it would be hard to implement. The memory would be the more out there thing but even that would be a matter of retrieval.
- lumost 4y ago
  The proposed method would improve the model anytime the model didn’t know which api to use. This requires the model to not hallucinate knowledge and some form of incremental training.
  - famouswaffles 4y ago
    Not necessarily. Even in people, memory is (albeit very complex) retrieval. I mean, how many of your x years of memories are in the forefront on your mind at any given moment? Not the creator so i won't make any definite statements on how it works on the memory side but if you embedded and stored the chain of thought "how to use x API" generations and then retrieved them anytime the model was going to use the same API, it would certainly simulate a memory of sorts without needing any incremental training.
    - lumost 4y ago
      Certainly possible! As with all things ml the question is how well this works. The author made up a scientific sounding claim without any backing data.

by lumost 4y ago

singularity2001 4y ago

I tried something similar with terminal copilot yesterday. The approach works for bash calls, why not for web calls?

famouswaffles 4y ago

GPT3 has demonstrably been able to run/execute execute and make API requests with python for quite some time now. You can test it too. See here https://twitter.com/sergeykarayev/status/1569377881440276481 https://twitter.com/sergeykarayev/status/1569377881440276481 https://twitter.com/sergeykarayev/status/1570868002954055682 https://twitter.com/sergeykarayev/status/1570868002954055682 None of this seems like it would be hard to implement. The memory would be the more out there thing but even that would be a matter of retrieval.

lumost 4y ago
The proposed method would improve the model anytime the model didn’t know which api to use. This requires the model to not hallucinate knowledge and some form of incremental training.
- famouswaffles 4y ago
  Not necessarily. Even in people, memory is (albeit very complex) retrieval. I mean, how many of your x years of memories are in the forefront on your mind at any given moment? Not the creator so i won't make any definite statements on how it works on the memory side but if you embedded and stored the chain of thought "how to use x API" generations and then retrieved them anytime the model was going to use the same API, it would certainly simulate a memory of sorts without needing any incremental training.
  - lumost 4y ago
    Certainly possible! As with all things ml the question is how well this works. The author made up a scientific sounding claim without any backing data.