Toolformer: Language models can teach themselves to use tools

Re:Toolformer: Language models can teach themselves to use t... by nadermx on Friday February 10, 2023 @12:14PM (#34742551)

I think this is probably the sort of way AGI will work.

[ Parent | on HN ]

Re:Toolformer: Language models can teach themselves to use t... by valine on Friday February 10, 2023 @08:54PM (#34748755)

A multimodal LLM using tools definitely feels like the closest we’ve ever been to AGI.

[ Parent | on HN ]

Re:Toolformer: Language models can teach themselves to use t... by oliverbennett on Sunday February 12, 2023 @04:26AM (#34761384)

If we’re at a place a LLM can make and use tools I’m excited about the limits of its ability to create tools... specifically to create new tools.

For me that kind of entailment is one of the more exciting pillars of general intelligence and unique among humans (I think?).

[ Parent | on HN ]

Re:Toolformer: Language models can teach themselves to use t... by PartiallyTyped on Sunday February 12, 2023 @06:28AM (#34762019)

The next step could be pairing multiple LLMs with code synthesisers. The LLM could describe the target API, a second LLM would build said API, and resolve the query.

I think at that point we will be very close to an AGI.

[ Parent | on HN ]

[dead]

Re:Toolformer: Language models can teach themselves to use t... by frozencell on Monday February 13, 2023 @07:35PM (#34783362)

What will the people complain about at the end?

[ Parent | on HN ]