AI Hustle: News on Open AI, ChatGPT, Midjourney, NVIDIA, Anthropic, Open Source LLMs: Google's Bard AI: Accessing Your Gmail, Drive, Maps, and Beyond

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.

Today, Google is evaluating the capabilities of its generative AI chatbot, Bard, and it

is setting it on a collision course essentially in my opinion with competitors like ChatGPT

so this upgraded Bard really comes with features that enable users to verify its responses.

It's also able to collaborate in real time, think Google Docs, right?

And most intriguingly, it's able to integrate with Google suite of applications and services

like Gmail, Docs, Drive, Maps, YouTube and even Google Flights and Hotels.

This significant update includes what Google is calling Bard extensions, though for now

these are only available in English.

Basically this was introduced at Google I.O., which is the rollout of this whole extension

was deliberately measured, but today Google has emphasized that it intends to offer these

features in a manner that prioritizes user safety and trustworthiness.

So Jack Krasowski, who is the product lead for Bard, elaborated saying quote, we want

to make sure that the way that we bring this to users is extremely rooted in three principles

that we have as it relates to the trust that we build with people that use Bard, which

is around transparency, choice and control.

So essentially users can opt into these features and can just as easily revoke permission.

So this isn't something that's just like default going to be on everything.

But in the, you know, kind of backdrop of privacy concerns, Google is being really upfront

about how it handles user data.

For instance, when Bard interacts with, you know, Gmail, Google Drive and Docs, this

is something a lot of people are concerned about.

They're like, I don't want to integrate Bard or AI into all of my data.

They're like essentially taking my Google Drive, taking all my docs, taking all my Gmail,

you know, companies have, you know, you know, classified information, intellectual property,

all sorts of things.

They don't want this integrated.

So a lot of people are worried about the privacy side.

But essentially how Bard is now being set up with this is that when it's integrating

with Gmail, Google Drive, Docs, it doesn't use that information for reinforcement learning

at all. So the policy helps maintain the, I think it's a really critical trust factor

they have to have to have here.

So Krasowski also said that Bard is not going to have access to store an entire Gmail inbox.

For example, it operates on a per prompt basis, pulling specific information as directed by

the user. So Google also assures that no human reviewers will gain access to emails Bard

interacts with. This I think is really important because when ChatGPT came out, like ChatGPT,

all of the things you tell it are going to human reviewers, right?

They have people that are reviewing the responses that they're getting for trust and safety

reasons and also for fine tuning and all sorts of other things.

And a lot of people were very eerie or weary of that.

I think, you know, you saw JP Morgan Chase, Apple, right at the beginning ban their employees

from using these tools because they didn't want, you know, things to leak out through them.

And so I think this is a really big move by Google.

I like, I don't know.

I don't know like how much confidential data I have, but I would not want, you know, to be

using this tool if I know that someone's sitting there on the other end reading my emails.

That feels like a massive, that's pretty much like the biggest invasion of privacy.

No one likes that.

So I think that that's a really good move on Google's part, making sure that doesn't happen.

The approach with to this really kind of aligns with Google's longstanding practice

around data security and privacy.

They said, quote, it's similar to how we've approached spam filtering and Google

services in the past.

Your personal information isn't read because we believe that that trust is the most critical

pillar upon which we built.

That was Krasowski again.

So I think when the Bard extensions are in play, the chatbot can perform a whole bunch

of different tasks like summarizing important emails or even planning a trip across by

essentially cross referencing Gmail conversations and real time flight and hotel data.

However, its capabilities do have limitations which Google acknowledges.

So for instance, Bard may struggle with identifying specific content like wine club

membership in your email, unless those exact terms are used.

So this is kind of interesting, right?

It's not like it can see all the data around it, get context and know where to look for

something unless you have like literal keywords attached to the data.

So this is interesting.

This is a big problem with like AI in general is that it's really hard to label all of the

data. Data labeling is a real thing.

And so they don't have a magical bullet for this yet.

But this is definitely possible if you use the right keyword.

So additionally, the new Bard can leverage multiple Google services in a single interaction,

which I think is really cool.

So if you're planning a trip with friends, Bard could, you know, pull the most convenient

dates from a Gmail thread, search for flights and hotels, you know, map your route to the

airport and even suggest activities at your destination through YouTube videos and do

it all in like one conversation.

So for features not utilizing personal data like YouTube, flights, hotels, maps, users

are opted in by default.

So be aware of that.

But they can easily opt out, apparently.

So the end game Google says is to extend this functionality to third party services.

But before that, the company wants to refine the features within its own ecosystem of apps

and services and really kind of enhance Bard's accountability.

So Google has added the Google it button to cross verify the chat's answers, tapping

the button validates every statement Bard makes against Google search.

So this is also very interesting where you can essentially, you know, verify if something

is actually true.

And this is allowing users to kind of further explore a topic.

And if Bard is uncertain about a piece of information, it's going to highlight it in

orange offering a visual cue that helps users discern the chatbot's accuracy.

This is also very interesting when it comes to, you know, AI and hallucinations and whatnot.

Being able to highlight things that's uncertain about is definitely I think a really big,

a really big feature.

So Krasowski is really optimistic about the future and said, quote, we are pretty excited

about taking this step towards building trust with language models.

Of course, we certainly want to be transparent when we're not confident or even when we make

a mistake and quote.

So I think this feedback mechanism is going to serve as kind of a basis for ongoing improvement

of the AI model.

Lastly, the update offers I think a really collaborative angle so users can now share

a Bard conversation via public link.

I think Chad's been doing this for quite a long time.

But this is essentially allowing others to continue the discussion and ask a chatbot

further questions on the same topic as part of its kind of global reach strategy.

Google is also extending Bard's existing English language features to over 40 new languages.

Overall, Google Bard is shaping up as I think a really powerful tool in the world of AI

driven chatbots.

They're making this thing a lot more robust integrating their ecosystem.

And I think this is really Google's strength, right?

They have a lot of services that are incredibly popular weaving them together into their AI

model could give them a bit of a competitive advantage over, you know, open AI, for example,

because currently Bard is useful.

It's not that bad, but I just feel like chatGPD always gives me better responses.

That's where my like default I always go to.

But if they were integrating into my own data, that does serve another use case of, you know,

kind of getting this AI experience based off of my own data, my own personal case and preferences

on things.

So I do see that there's a lot of value and perhaps Bard could help kind of pull itself

ahead in a race that it seems to be lagging.

It's not definitely not in first place.

And so I think this would be really interesting to see how this rolls out and how this is

adopted by users and what the reception is.

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

to join our chatGPD 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 chatGPD.

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.