AI Hustle: News on Open AI, ChatGPT, Midjourney, NVIDIA, Anthropic, Open Source LLMs: Meta's AI Boost: Image Magic & Chatbots on IG, Messenger, WhatsApp

Welcome to the OpenAI podcast, the podcast that opens up the world of AI in a quick and

concise manner.

Tune in daily to hear the latest news and breakthroughs in the rapidly evolving world

of artificial intelligence.

If you've been following the podcast for a while, you'll know that over the last six

months I've been working on a stealth AI startup.

Of the hundreds of projects I've covered, this is the one that I believe has the greatest

potential.

So today I'm excited to announce AIBOX.

AIBOX is a no-code AI app building platform paired with the App Store for AI that lets

you monetize your AI tools.

The platform lets you build apps by linking together AI models like chatGPT, mid-journey

and 11 labs, eventually will integrate with software like Gmail, Trello and Salesforce

so you can use AI to automate every function in your organization.

To get notified when we launch and be one of the first to build on the platform, you

can join the wait list at AIBOX.AI, the link is in the show notes.

We are currently raising a seed round of funding.

If you're an investor that is focused on disruptive tech, I'd love to tell you more

about the platform.

You can reach out to me at jaden at AIBOX.AI, I'll leave that email in the show notes.

So the first thing I want to talk about here is the fact that Meta is coming up with several

new AI driven bots across its suite of messaging apps.

So that's including WhatsApp, Messenger and Instagram.

And this new offering, which is currently in beta and it's available to select users

in the US including essentially has an AI assistant named Meta AI that comes really close to mimicking

existing chatbots like chatGPT and Anthropics Clawed2.

This is very, very interesting.

Essentially they're building into your personal text messaging or whatever, chatGPT.

Now this is interesting because of course we've seen something very similar out of Snapchat

who has my AI, but that's just, you know, that's directly an API on top of OpenAI.

So I'm inclined to be a little bit less impressed with what Snapchat's doing because it looks

like Facebook here is, and you know, Meta is doing something completely new with their

own technology, which is really impressive.

So all of this was unveiled today at, or yesterday actually at Meta's annual connector developer

conference.

And Meta AI is designed to essentially assist in planning trips, answer general knowledge

questions, even searching the web.

And this is going to be really interesting because chatGPT took this feature away, the

web search, then they just added it back a few days ago.

But here we have Meta doing it and, you know, if you're wondering how they're going to do

this whole web search thing, they're actually partnering with Microsoft's Bing for real-time

answers.

This is kind of interesting because like on the one hand people are like, what, I thought,

you know, Microsoft and Bing had this special partnership with OpenAI because they invested

like $10 billion into them.

But it's interesting because Microsoft in that investment, I think subsequently has

kind of said like, look, we made this big investment, we made this big play, we hope

they're the winner, but there's a very good likelihood that other tools emerge the winner

in this space other than chatGPT.

So it looks like they're not really tying themselves down to one specific partnership

they're hoping to win.

They're kind of playing in a lot of different fields and allowing their technology to be

used for a lot of different reasons.

And of course, a big part of their whole business model is increasing usage of Microsoft's Bing,

you know, they're kind of fighting with Google over there.

And so I think for them, they're kind of excited to have another avenue that's going to generate,

you know, more traffic for Bing specifically.

And it's kind of interesting because I just feel like the way the tech of landscape is,

I would be surprised if I saw like Zuckerberg and MetaGo do like a partnership with Google

or something, especially with, you know, I don't know, Bard and some other things.

So very interesting, but Bing is now being integrated into this whole thing.

And CEO Mark Zuckerberg revealed that MetaAI is slated for integration into Meta's recently

announced Quest 3 VR headset.

So that's going to be another thing that is happening.

And this whole new AI assistant appears to be Meta's second attempt at a virtual assistant.

So this is following the somewhat failed experiment with M, which was a limited release, partially

human aided virtual assistant from like years ago.

So M was only ever available to about 2000 people living in California, and it suffered

from development neglect, I think, never actually really got off the ground or scaled in any

way.

However, MetaAI seems poised for a wider role.

I think obviously one of the big problems with M that they had in the past was, you

know, the fact that it was like partially a human virtual assistant answering questions.

So now that you actually can automate this whole thing and roll it out to hundreds of

millions of people, I think, obviously, it's a lot more viable of a solution to a platform

like Meta with billions of users.

So unlike its predecessor M, which involved human intervention for completing tasks, MetaAI

is completely automated, and it's running on a custom made large language model.

As we know, you know, Meta is quite quite a pioneer in the LLM space, they've been working

on this for a long time, they've done a lot of open source projects.

And so I think the model has been trained to give very concise answers and can even

refer back to previous conversations.

So Meta didn't just stop at MetaAI.

Also yesterday, the company announced the rollout of a series of AI characters.

So these are chatbots designed to mimic personalities and celebrities.

Now, if you know me, you've heard me talk a little bit about character AI.

And this is like, it's kind of funny because by web traffic character AI, or I think maybe

mobile traffic, it's like the second most used chatbot tool behind OpenAI's chatGPT

as far as iOS downloads in the US.

And I just think it's super funny, essentially, like character AI lets you make different

like personalities and you can chat with them, right, like Leonardo da Vinci or Einstein

or something.

Meta is doing a little bit different, and I think they're going more for like modern

day personalities.

So it's designed to mimic people like Kendall Jenner, Dwayne Wade, Mr. Beast, Paris Hilton,

Charlie D'Amelio, and Snoop Dogg with a bunch more to come, and these chatbots compete

or complete with, you know, have a bunch of profile images that can either be illustrated

or real life, and they're going to stay on Meta's messaging platforms.

So interestingly, as users engage with these avatars, the characters will animate in response

to the conversation.

This is going to be very interesting to see how that works.

I believe I even saw a news article that said for this whole project, they had like

Mr. Beast come into the studio and like film video of him because you know, they're doing

these like animations and stuff.

So very, very interesting, and obviously these are actually in collaboration and partnership

with these influencers, unlike, you know, something like character AI, I think it's just people

that are making kind of these like person, these like, you know, they're making these

fake people that you can chat with.

So very, very interesting.

I think while the potential application for Meta's new chatbots are quite interesting,

some people say they don't come without challenges.

Meta revealed that it had spent 6,000 hours identifying problematic use cases and testing

the models internally before making them public.

And the company also claims to have developed technology for flagging content that it goes

against its policies and has restricted web searching capabilities to only Meta AI.

So furthermore, Meta has opened its new chatbot features to researchers through a bug bounty

program.

This is something that Meta does a bunch that I do think is good, like I think they're quite

aware of like the hacker community and whatnot.

So I think it's good they focus on that.

On the business front, Meta launched AI Studio, which is a toolkit to help companies build

their AI chatbots for Facebook, Instagram and Messenger.

So initially in Alpha AI Studio, this is the second big thing that was announced here.

It's called AI Studio, and it aims to let brands create bots that reflect their values

and improve customer service experiences.

So the primary use case is kind of like that they're really targeting our e-commerce and

customer support.

So it's very interesting.

So support these kind of business focused endeavors, Meta plans to release a sandbox

tool allowing anyone to dabble in creating their own AI.

So the sandbox will eventually be integrated into Meta's metaverse platforms like Horizon

Worlds for powering NPCs across various metaverse experiences.

So although Meta is...

And okay, one thing I do want to note on that that I think is very interesting is platforms

like Steam right now have a policy where they don't want any AI on their platform.

And I think maybe some of it's for copyright issues, I'm not sure.

But in any case, I've seen a whole bunch of news stories about people trying to use AI

to make these NPCs, these non-player characters, it's like the guy behind the bar table that

you ask a question and he tells you to go a place in the video game.

Steam if you try to use AI or chat GBT to come up with just like a conversation from

this guy, Steam was like banning people's entire projects, not even just like not allowing

them to push an update but just banning them because it wasn't allowing AI.

So what's really interesting to me here is that Meta on focusing on their metaverse and

some of the games there is specifically like trying to build this technology in which I

think a lot of people want to see.

So it's going to be very interesting to see how that kind of rolls out.

Although Meta is optimistic about its AI capabilities, there are definitely some challenges.

I think the large language models powering these chatbots have been known to kind of

exhibit biases and occasionally refer to inaccurate things or fabricate information with Meta

AI and AI Studio.

So the company seems to be really focused on advancing the field but is also going to

need to, it's also going to have a pretty high bar of public and technical scrutiny.

So I think as these chatbots services roll out, the real tests will be how well they

adapt to and meet the complex needs and expectations of users and we're going to kind of look at

something similar to what happened with Snapchat to see how this rolls out.

But I think given Meta's kind of promise for a broader release, I don't think we're going

to have to wait long to find out.

And I did see a demo of Mark Zuckerberg talking about this.

It was really interesting.

He specifically said like, you're not going to have to just use like a specific chat like

on Snapchat.

Like the My AI friend or whatever thing and you can chat with AI there.

This isn't going to be just that on Meta's platforms.

You can actually invoke or summon the AI like in a conversation with a friend.

If you use like the backslash imagine, which is kind of funny because they just stole that

straight from mid-journey, like whatever.

It's actually able to generate images so they have their own image generator, which is going

to be really cool.

And like so you could be like literally in a conversation with a friend.

They say something funny.

I don't know.

Like a meme or whatever.

And so then you just say like, I'm going to make a meme of like a surfboard on top of

a car flying through the sky, going down a wave and you know, you say imagine first

and then Meta will be able to generate that image right there.

So you kind of custom images.

I think it's going to be kind of fun to be honest in conversations to be able to do that.

Like imagine mid-journey was just like built in so to your conversations with your friends

because Meta kind of has a monopoly on the whole like, you know, chat thing when they

got like, you know, Instagram DMs, Facebook Messenger and WhatsApp.

They really have a huge part of the chat market.

So I think this is actually going to be way bigger than something like Snapchat.

In addition to this, I think you can invoke the AI in conversations with your friends.

So if someone says something and you want to like, I don't know, you want to be that

annoying person that has to fact check them right there, you can summon the AI to settle

a debate or whatever.

In any case, I think it's going to be really interesting.

And honestly, I do think this is a fairly relevant AI release.

It's a big use case.

A lot of people are going to be working with AI here.

And I think what we're going to see is in the next, you know, year or so, AI is going

to be so deeply integrated into everything we do.

It's going, this is a completely new revolution that is about to change everything.

So very excited to continue to follow.

If you are looking for an innovative and creative community of people using chat GPT, you need

to join our chat GPT creators community.

I'll drop a link in the description to this podcast.

We'd love to see you there where we share tips and tricks of what is working in chat

GPT.

It's a lot easier than a podcast as you can see screenshots, you can share and comment

on things that are currently working.

So if this sounds interesting to you, check out the link in the comment.

We'd love to have you in the community.

Thanks for joining me on the open AI podcast.

It would mean the world to me if you would rate this podcast wherever you listen to your

podcasts and I'll see you tomorrow.

Machine-generated transcript that may contain inaccuracies.