Computational Conversations EP#1 | How do Al, machine learning and LLMs relate?
HTML-код
- Опубликовано: 9 июн 2024
- Conrad Wolfram and Jon McLoone explore the relationship between AI, machine learning, LLMs and computation. What's new, what's hype and what's possible?
Развлечения
Very much enjoyed this!
Interesting -- hope that there are more like this.
I cannot find "Computational Conversations" in my podcast app. Are you available as a podcast yet?
Good stuff. Looking forward to future episodes. Will the be a dedicated podcast RSS feed for Computational Conversations episodes?
never thought about AI being thought about like that. I do think that AI is just I (intelligence) but intelligence is itself always shifting and not grounded in any objective basis. If we do break it down into wolfram-esque ideas, then intelligence is at most just a system running rules and intelligence no longer holds any qualitative meaning for the purpose of making comparisons. So in this sense ya, he's right in that AI as a naming scheme is meaningless but its just a bit deeper than just shifting of intelligence gap.
Machine Learning and LLM's are two such type of systems (programs) that follow rules to produce some output, and you could just consider them as intelligences themselves, under the above definition.
What do you think?
Sudomerge: rrr4:(r:0)
+Minsky
Her name is said thurnberry but is spelled like Daniel from ell. Check: Nader+goff+ellsberg:y
!!-
Whaaaaaaaaaaaat
can u please use ai to fix mathematica 'abort evaluation' which never works w big data in memory - it just hangs - filed n tickets w/ tech support since 2015, no material response from support. analogous op works in python + jupyterlab >> keep talking sht
My octopus teacher. Consciousness ed. Link with octopus circuit. Find. Link with joules thief circuit. Find. See vi hart, Kantor and Hilbert for nested infinity memory array dimensioning primer. Bring forward integrating operations. Stop thinking of yourself as 1 and running out of numbers.
The system is error-correcting in a way that amazes Neil degrasse Tyson. You possibly remember. Watch the Sagan version. Sorry to say it was much better than the stuff of Neil's I managed to illicitly acquire in the UK.
Call the piss off lice.
Hugh Everett can be verified accurate. Because he was accurate. We do not exist in a universe and your attempt to describe time function was something I thought I might never hear, condemned to the BBC's DReemer Brian Cox batting forever how blooming vast, vaaaast it all is I tellz ya.
Discovering the features and how to use them of the system that can't say anything other than it makes life because it enjoys doing so ..... worth doing.