Lex Clips
Lex Fridman Podcast full episode: https://www.youtube.com/watch?v=nAMjv0NAESM
Please support this podcast by checking out our sponsors:
– SimpliSafe: https://simplisafe.com/lex and use code LEX to get a free security camera
– Eight Sleep: https://www.eightsleep.com/lex and use code LEX to get $200 off
– ExpressVPN: https://expressvpn.com/lexpod and use code LexPod to get 3 months free
– BetterHelp: https://betterhelp.com/lex to get 10% off
PODCAST INFO:
Podcast website: https://lexfridman.com/podcast
Apple Podcasts: https://apple.co/2lwqZIr
Spotify: https://spoti.fi/2nEwCF8
RSS: https://lexfridman.com/feed/podcast/
Full episodes playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOdP_8GztsuKi9nrraNbKKp4
Clips playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOeciFP3CBCIEElOJeitOr41
CONNECT:
– Subscribe to this YouTube channel
– Twitter: https://twitter.com/lexfridman
– LinkedIn: https://www.linkedin.com/in/lexfridman
– Facebook: https://www.facebook.com/LexFridmanPage
– Instagram: https://www.instagram.com/lexfridman
– Medium: https://medium.com/@lexfridman
– Support on Patreon: https://www.patreon.com/lexfridman
Right now it is more like using intuition for everything so the next step would be to add logic and critical analysis
What we need is GPT-Science!, feed it all the scientefic papers and math equations, imagine if it will be able to build new unimaginable papers and its own new math equations just from the question.
keep chasing this fallacy that a computer can be human. zzzzzzzzzzz…….
The conversational Hamiltonian
AI is the equivalent of the Hubble Telescope. It’s going to help us recognize how small we are.
Quote, " GPT-3 is a major advance in AI ". No it's not, as there's already AI in computers, period. All you're, or anyone is doing at this point, is way enabling AI over humans. This act is sheer self suicide.
Is all cognition/reasoning like that? Quite obviously it's not, given that very few humans have read "the whole Internet" and we can still imagine things and put together reasonable sentences fairly easily.
It's not conscious. But it doesn't need to be in order to output intelligence.
I asked GPT-3 what Scott was going to say next.
Right.
No I DONT KNOW !
STOP SAYING Y O U. K N O W !!
I write idea for videogames and my question is: is it possible to ask this ai to have a look of what I have written or you cannot have this kind of option, as it would be amazing if I could ask to read my work and then see what the AI is able to come up with.
Can it find independent energy if it were threatened with cutting off its energy source?
who is the guesser on the left?
Skrew gpt3 it serves no purpose
I can't listen to this guy.
it would be a great idea if IBM could make a new brain for Watson to get it to be smart enough to replace doctors so we will have more time to get to know our doctors on so they can figure out what is wrong with us and how can we be cured or get better faster .
UN, World Bank document, and installers prove Covid 19 is a chip used in 5g. 1=A,9=I,=AI.. 5g has Laser capacity! Ask Jesus to forgive your Sins, and come into your life from your heart, Now, Miss Hell !
Ex Machina
Wow, finally someone discussing AI without being ridiculous about it.
If he can go 60 minutes without mentioning politics then he should be a new-media superstar.
These first few iterations seem to be more of a proof of concept. Obviously, GPT is a work in progress; it will be fascinating to see what the later versions can do.
Do you really need a BIGGER GPT?
Is the social part not what's missing? What will consolidate certain habits in 'thought' and the reproduction of the reasoning of others in order to 'align' and 'be together'.
Let a million GPT-3's communicate freely and see if they agree and what they come up with.
(war?)
This guy loves to talk. Legit doesn't want to listen or learn more, just wants to sound like he created GPT-3
https://youtu.be/G_-BBniFFCM?t=303 What is the Mount Everest Question? Being able to reason whether an object is bigger or smaller than Mount Everest?
If this guy says "you know" one more time….
When we are speaking and writing we aren't just predicting the next most likely word, our brain can alternate between different complementary subjects or specify and develop certain ideas. GPT-3 and other NLPs are just minimizing distances between vectors representing words, they just predict the next most likely sequence, nothing comparable to thinking or reasoning. Most of the GPT-3 texts I have seen are perfectly written (sound human) but are just random nonsense. The model learned from the random gibberish of the internet it will output also random gibberish.
You know you know
YOU KNOW……………..
I just look at the video "studio" quality and can see the question as "where were you on the night of the 3rd at 11:35pm"…
GPT-3 already does "reasoning" by my understanding of the definition of the word. It can apply its learning to solve problems it has never seen before. GPT-3 CAN balance parenthesis well enough to write working code. GPT-3 DOES have some visual reasoning and can draw very simple pictures and create UI layouts for applications.
https://youtu.be/PqbB07n_uQ4
This makes me so happy. Thank you. I saw Eric Elliot's conversation with it. Amazing
Sounds and moves like Ben Goertzel
George Costanza is looking better these days.
This interview seems very unprofessional and difficult to listen to.
I once set out to create a general word tree built from an input of only coherent and factual statements, simply tracking the ordering relationships of the words. That alone seemed enough to create a network of knowledge that could be queried in lots of interesting ways. The key for GPT will be to add more co-agents to apply the knowledge network in meaningful ways and as part of that build improved knowledge networks that work better for the problems at hand. As you deploy more agents that use and refine their knowledge you will get emergent general AI out of it, and possibly even self-aware agents.