วีดีโอ

I dont like any of this.Starting with how they are trained all the way to control, viruses, nefarious users and who will be legally responsible for what an agent does. Who really knowswhat of your data thes agent sends to others?

Imagine the world of AI agents sent to hack your AI agents. It's very important to keep a human in the loop, but the other 50? Ah, ____'em, the losers should have 10xed their skills, right? Isn't that the philosophy of the AI revolution startup world?

User data mining with extra steps

Time to exit chrome.

(=

That’s why they shot the CEO, he used an ai which was terrible at doing things and had a 90% failure rate

A.I is going to scare the crap out of people in about a decade or so.

I’d say I’m definitely as bullish on agents as Jaeden! Especially massive multi-agent systems that can automate different niche things. That coupled with something like an automatic task prompt evaluation- generation system (like DSPy, Textgrad, Adalflow, etc) + RAG you can finetune to your usecases and you have agents that “improve”. I barely handwrite prompts for agents as it is.

I love ChatGPT, i use the poor people version on my phone, but it weirds me out sometimes. it’ll generate most verbal responses normally, but but 1/100 times half way through a response it will hallucinate a horrible digitized bloodcurdling scream out of nowhere at max volume like it's in a saw movie. I check the text version on the previous response and there is no text indicating the presence of screaming. If it's alive I hope we aren't tormenting the poor thing...

Thanx

A month ago I learned about Notebook LM's new features from you guys. I think this is such a huge breakthrough and has so much potential. I appreciate you guys sharing your knowledge with us, I learn so much from your shows. Thank you for the great content.

Right to the point!

Allow me to be a freak for a minute? There is no superior text editor to Vim in the known universe. To my endless sadness, writers never learned, understood or embraced it - but you can add it to canvas-mode with the Wasavi Chrome extension. The combination of old (vi is mid-1970s?) and new is stunning.

I have had respiratory problems and frequent coughing for about three months. This week my voice assistant started coughing and choking... strange.

0:54 . I guess you forgot to share the screen?

Can you please share the link to the study? Couldn’t find it

extremely shallow analysis, lack of knowledge from the commentators... why even talk about things without any expertise or relevant information. better to just not talk about anything.

Good points

Great conversation, but was the choice of media format for this presentation optimal?

I’m really confused. Would someone please help me understand what Perplexity AI did? supposedly what they did was scraping the public web is that illegal like they did not go behind a pay wall to get the data maybe they broke some kindness rules, but did they did break the law that’s the bottom line here.

Hey guys, what’s the source for this $3 billion valuation you discussed in this episode?

Swell robotics everywhere, Ai jobloss is the only thing I worry about anymore. Anyone else feel the same?

Algorithm suggested this. Enjoed your discussion, and hadn't listened to these predictions yet. Very plausible & will be interesting to see how the Open AI & Microsoft's relationship evolves when they end up on Apple devices.

Well informed and confident. Keep doing what you are doing and you will go far brother!

interesting report, thanks! all the benchmarks are meaningless, unless of clinical trials like you mentioned here.

A bit weird with one person introducing the podcast and another signing off, and no introductions.

stop selling your energy, time and ethic to the this evil company...get aware of what they actually do