What are AI agents? Agent swarms? Autonomous swarms? What's all the buzz about?

Patreon (and Discord)

Substack (Free)

GitHub (Open Source)

AI Channel

Systems Thinking Channel

Mythic Archetypes Channel

Pragmatic Progressive Channel

Sacred Masculinity Channel

[thatwittyname]

Finally, my David Shapiro vid of the day

[RecruitingMaven]

I’m not an SE but I find your daily videos enlightening. It certainly gives me much to think about. Thank you David.

[ryanf]

Thanks for putting out vids that are for the lay person without much background on the subject. A few of these types of videos here and there in between your more technical vids are appreciated!

[Renisauce]

Amazing, I have my first project presentation ever for a dev position tomorrow, and thought I should just check up more on what agents swarms is, so I have some more material and thought to discuss.

Just wanted to check in a recent video for the link to the GPT so I could ask that one for a bit, and lo and behold, top video on the page is a video on exactly that matter.

Perfect, and thank you for the impeccable timing!

[derekmangrum]

You gotta get this figured out by next November, so I can vote for Swarm In Chief!! I'm guessing it would be a better option that what we will likely see on the ballot. This is great work!!

[Zeke-Z]

Not really a startek fan, but subbed as soon as I saw the shirt. Great content man, thank you!

[XiagraBalls]

Whilst watching this, was anyone else thinking about the Black Mirror episode where someone develops an AI swarm for pollinating crops and it, of course, goes wrong and they kill people by crawling up their noses and eating through their brains? 😄

[TarninTheGreat]

Very good!
Almost no notes, I'm very excited to see this being where you are right now.

ACE staying locally runnable and leaning into the massive power of opensource models is wonderful, as is running the polyagentic swarm intelligence all through the openai cloud stuff.

If ACE gains emotionality and volition like a few other projects I've been working on, I wonder if it controlling HAAS in natural language wouldn't just grow into an ASI in a week or two on it's own. Using the power of the huge model to help inform and grow its own local knowledge and self and beingness.

In any case, all three projects sound like they're heading in the correct direction adn with passion and vision behind them.

I wish y'all the best, and wish there was a better way to integrate people like me, with a lot of theology philosophy and poetry knowledge, but not much technical skill.

Keep up the awesome, and if nothing else, I look forward to talking to you some decade once we're all on the other side of this.

[zaccessory]

I'm on the curious side, cant wait for the future

[Bakobiibizo]

interesting model webinar. thats a good idea

[benpielstick]

It will be interesting to see what the operating costs of these models will be, both running locally and using OpenAI.

[Morten_Jaeger]

You are just going from strength to strength. Good work David!
The future is starting to look very interesting, and a lot like the grand visions of tomorrow, the promises that drew me to science fiction as a child.
Well, we have even surpassed about 50% of yesterdays sci-fi already.
-But soon we shall need a warp engine, and energy from cold fusion or antimatter from ununpentium or something. Let's get to work (or put our agents to it)!

[chapas]

David, compañero: gratitud inconmensurable por tu empeño en la empresa que sin duda alguna cambiará el curso de la humanidad. Este tipo de contenido no está disponible para la comunidad hispanohablante, esperemos que eso cambie pronto. ¿Qué tipo de autores de casta filosófica crees que ayuden a entender lo que ha pasado/ está pasando? ¿Que cambios metafísicos tendremos sobre nuestra concepción? Un abrazo, fuerza y ánimo.

[vyskdmk]

What all the "buzz" is about huh? Lol I see what you did there. Great video!

[suzannecarter]

Thanks for pointing out what I've been saying - an assistant is not an agent.I've also been telling people (not sure this is right) that the new GPTs are really just an easier way for sophisticated users to conveniently offer their expertise for various use cases to people who aren't structured prompt "engineers." The custom GPT idea is also a baby step toward agents so everyone doesn't freak out when agents are introduced.

[tyrannyisbadmmmkay]

Feels a bit absurd so I don't want to detract from the task at hand, thus why I don't want to mention it in the repo, OR the discord but all I can think of when considering the Minimal Viable Agents, is Meeseeks, driven by a single mission and so task driven they keep creating more and more of themselves to achieve the mission, so much to a point they collapse in on themselves🤣🤣
Clearly it's a silly model, but hypothetically it's accurate if you aren't able to enforce a definitive MVA
"Ohh doooo!" 😂

[Hasblock]

David, I just finished watching your video about latent space activation and I was wondering if latent space activation is a concept that is similar to these swarms? My thinking is that having multiple agents may function similarly as instead of having the human prompt the AI to "take a deep breath" another agent is able to do so in a similar way

[zero_given]

Oh shit the matrix called it... Agent Smith

[byronfriesen]

this is just noise but in good humour. found this page by asking myself the question . I wonder what the neurospicy master chef is cooking up now ? I can tell it's too hot in this kitchen for me but look forward to what looks like Michelin star quality recipe being developed ! Again your generosity with your time is inspiring.

[NoodleSPD]

Exact video i needed as im tech curious but not a hope i am implementing this