Meta Replaces ChatGPT

This is Synthetic Mind, the A.I newsletter thats more iconic than Vin Diesel in a car movie except we take 5 min, not 12 movies.

Its finally Friday. Heres how were closing out the week:

  • Apples Secret A.I Plans

  • Meta Replaces ChatGPT

  • OpenAI Holds Back Powerful A.I Tool

Read Time: 5 min

Apples Secret A.I Plans

Lesson of the day: Dont believe everything you hear

Apples promised us over and over again they were NOT working on anything A.I related other than a few minor updates.

But yesterday Apple formed a leak, and boy did they let it pour

The secret info was so juicy the Apple stock shot up seconds after the report was released.

When I heard what it was, everything began to make sense - Like why Apple banned their employees from using ChatGPT at work

The secret?

Apple is working on an A.I chat of their own called AppleGPT

(missed opportunity for an iChat but whatever)

We dont know too much about it. Apple is acting like nothing has happened

But heres what we DO know:

  • Its going to run on an LLM called Ajax. This is like the GPT-4 to ChatGPT.

  • Ajax is being tested in some Apple products right now. There was talk a while back about some A.I updates with Apple TVs but alls quiet on that front atm.

  • Apple will release an official announcement within the next year. Yeah, weve heard that one before

Some are guessing AppleGPT will become a combo with Siri, and others are hoping for an AppleGPT iPhone assistant.

The bottom line is we dont know what Apple has planned, but they better start now if they have any hope of catching up to other companies.

The Ultimate A.I Community

Lets be real for a second

A.I is f*cking confusing.

But we make it easier with Profit Snack.

Profit Snack is our tight-knit community of 50 driven and successful A.I enthusiasts you can learn from and are rooting for your success.

Find a business partner for your start-up, or hear about the next big thing in A.I.

Meta Replaces ChatGPT

When ChatGPT first came around in November 2022, Microsoft was first in line to get in on the A.I train.

But now, 8 months later, Microsoft is rethinking its plans

Why? A couple of reasons:

  • OpenAI is causing more drama than an elementary school playground. Theyre going to court, getting letters from the FTC, and fighting with the EU. It seems Microsoft thinks OpenAI costs more than its worth.

  • OpenAI is releasing two Microsoft competitors: An OpenAI Marketplace and an OpenAI Office Assistant.

  • Meta gave them a better deal. Now this is new.

Meta just broke the news that theyre releasing a new A.I chatbot named Llama 2 which will be available to everyone.. for free.

Why is this big?

RN businesses are thirsty for A.I tools, but not just any A.I tools

Businesses want A.I software they can train with their data and build new tools with (AKA open sourced).

With Llama 2, Meta is giving them just this. again, all for free.

Something ChatGPT has yet to offer.

So Microsoft hopped in early and made a deal

Microsoft will help fund and support Llama 2 in return for using the chat in Microsoft Azure -a spot originally promised to OpenAI.

I have to say, I think its smart for Microsoft to look elsewhere and have multiple A.I companies supporting their products.

Like we talked about yesterday, we will start to see a lot of A.I companies crash and burn as the A.I bubble grows and big tech giants monopolize on A.I resources.

OpenAI Holds Back Powerful A.I Tool

What started as a good intention turned bad fast as OpenAI tried to use ChatGPT to help the blind. How you ask?

One word: Hackers.

Lets start from the beginning.

A few months ago Open AI partnered with Be My Eyes to create a virtual eye using GPT-4 (what powers ChatGPT and Bing).

Think of it as a virtual eye that can read menus, help people cross the street, and most importantly of all. Use facial recognition to identify people.

At first, it was no big deal

Then two things hit them at once:

  • Countries and states started banning facial recognition and threatened some big a** fines for breaking the rules.

  • Hackers found a way to use the tool to bypass top-security systems and potentially track people via their faces popping up in the system.

Really not good.

Especially as the EU tries to pass their A.I regulation laws which will ban A.I facial recognition by tech companies.

Because of this, OpenAI and Microsoft plan to hold back the full version of the tool until they can figure out a way to solve this.

What'd you think of today's edition?

Login or Subscribe to participate in polls.

Join the conversation

or to participate.