Bold AI Predictions From Cohere Co-founder

[@Rockyzach88]

Totally agree with ~ 46:45. I got lucky that these things are coming out right as I'm finishing up my degree.

[@shay5025]

Ben Goertzel's talk the other day has really changed my opinion of the utility of LLM as the basis of any AI architecture. I'm simply a layman but wrapping everything around an LLM doesn't seem like it's going to get us very far.

[@MartinLaskowski]

13:02 I do think people hold LLMs to unreasonable standards. They are a wonder, not a panacea.

[@TheRyulord]

The people selling them have been marketing them as a panacea and it's entirely reasonable to hold the companies to the standards they set for themselves by making those claims.

[@hubrisnxs2013]

Love Cohere, thanks for the interview

[@hellojeezai]

As Aidan Gomez mentioned in a recent Politico interview, AI models will become more heterogeneous, with different models being used for different tasks. Ivan shares this view, predicting that learning and adaptation in AI models will be an ongoing process, as these systems continue to be updated and refined.

[@barberb]

prediction: the nobel peace price will be awarded to chatgpt

[@АмиЛаза-г1ю]

ridiculous

[@TheExodusLost]

@@АмиЛаза-г1юnot ridiculous. He’s talking GPT 5/6 by the way.

[@SjS_blue]

I mean the physics nobel went to the godfather, so why not

[@PhượngLongTrần-n9l]

Thanks

[@bobbyevans8496]

39:25 no flops are spent on backpropagation during inference 🤓

[@Aesthetic_Euclides]

Great conversation!! Another interesting way GenAI is impacting us is in how it lowers the bars for young entrepreneurs (literally teens!). All the steps needed for MVP are suddenly much more approachable (brainstorming, coding, automation, marketing, networking, social media, emails, summaries, note-taking, learning in general ...). You barely need to hire people and in the case that you do or don't, you will be much more productive. It's just like having books, calculators, excel sheets, internet... before they were available. A huge speed-up imo. I'm excited to see more and better tools and how people make use of them.

[@NabiruBogdan]

Nice

[@Dri_ver_]

Really funny to say this technology equals the playing field. I think the average worker would disagree. It's so crazy to continue to watch people talk about some kind of Star Trek utopia while forgetting that in Star Trek they actually don't have capitalism!

[@byt3blitz]

I agree! Just cause someone, with no technical experience, can tell a statistical model to produce some shitty todo app does not mean it equalizes the playing field. A competent developer using LLM’s, will run circles on someone with no experience, including this guys mom.

[@Lord_Liverpool]

0:18 It would be nice if they could tidy up all the nasty cables under the desks.

[@Walter5850]

When you have the ability to make a lot of money with a dangerous tech, I guess you'll go a long road to rationalize your actions.

[@Rockyzach88]

You can be for AI safety and not act like everyone who produces it is a murderer. I hate the people who are tunnel visioning it and think humanity needs to basically get out of their way and step aside for w/e VC funded AI shit they have, but then there are people who are coming with that same energy from the opposite side as well and it doesn't look good. Not necessarily saying that's you but I figure this is a good enough comment to throw this under.

[@angloland4539]

❤️🍓