🦾How You Can Leverage:

Are there some super weird AI experiments going on under our noses? 

Maybe. 

Maybe not. 

If you’re not dorks like us, maybe you’ve missed all the Project Strawberry 🍓and AI agent hoopla. (Trust us… it’s a lot!) 

So even though we’re big on bringing y’all facts and not diving into AI rumors and rumblings, we had to take the OpenAI/GPT-5/Project Strawberry buzz head on in today’s show. 

Here’s why: 

- Over the past 18 hours, there’s been a TON of (AI-generated?) movement

- Big publications, like Bloomberg, are reporting on recent developments

- OpenAI’s CEO Sam Altman has even been participating, which means this is more than rumors and will likely impact the future of work. 

Welp. 

Buckle up. 

Ready to go on a wild strawberry ride? 

Let’s goooooooooo 

As we always say, you might wanna give our whole show today a watch/listen for full facts. But here’s the skinny on what’s happening: 

A. Seeds planted

These Strawberry seeds actually started last year as Q*, which was widely reported on. Essentially, Q* was reported as a super advanced and (scarily?) capable model with human-like reasoning.

Reports actually tied the whole Sam Altman firing and the departure of OpenAI cofounder and chief scientific officer, Ilya Sutskever, to Q*. 

B. Q* is now Project Strawberry

The Q* project (Q-Star) is now reportedly called Project Strawberry. The super-secret project is reportedly tied to a new OpenAI model that has advanced reasoning capabilities to autonomously navigate the internet and perform deep research. 

C. The (Twitter) hype might be real. Or AI. 

The hype over the past week or two has soared, as Sam Altman himself has teased Project Strawberry online. 

Also, two mysterious Twitter uses have skyrocketed in visibility as they cryptically pump the tech, as many are guessing if they’re humans or if it’s actually future OpenAI autonomous tech running incognito. 

And giving these users more visibility, even Sam Altman has been conversing with them publicly about the 5 Stages of AGI. 

(And yes, even Bloomberg was trying to get a comment from Mr. Strawberry Guy (or Mr. Strawberry AI?), and the account declined. We’ve had some Twitter DM discussions, but nothing to report.) 

This is where the Strawberry gets spicy. 

On Thursday evening, the other mysterious Twitter user, Lily Ashwood, joined in on a Twitter/X spaces audio chat. 

Anywhos, MANY in the Space were quick to point out that Lily kinda sounded and responded like an AI. 

Take a listen:

(Sure, it could be a human trying to sound like an AI for likes/clout. We get that. But it did kinda sound AI-y to us and many others from today’s livestream who weighed in.) 

But y’all…. She/it/cyborg was responding in real-time to other users in the space.

If you didn’t know, the new Advanced Voice Mode that OpenAI had to put on hold because of safety concerns is technically capable to listen to and speak with humans in real-time. 

Oh.... The timing. 

So throw in the rise of Mr. Strawberry Guy/AI, Sam Altman tweeting Strawberries, all of these new reports, Lily Ashwood seeming like an AGI agent out in the wild…. WAIT. THERE’S MORE!

Within hours of Lily Ashwood kinda eery debut (sorry if you’re actually a human), a group of world-renowned researchers dropped a pretty big study. 

The topic? 

New methods to fight online AIs trying to be human. (Seriously. Go read it here.) 

In other words, some of the smartest academics in the world are kinda saying there’s gonna be AIs simulating real people online soon, and they’re warning us all it’s gonna be super hard to tell the difference. 

(These are super legit researchers…. OpenAI, Microsoft, Harvard, Oxford, MIT, etc.) 

Anyway, they suggested something called "personhood credentials" (PHCs) to combat AI-driven online deception.

PHCs are like digital ID cards that show you're a real person without revealing your identity.

Like…. Cmon. Is the timing of this all not 1% weird to y’all? 

In other words…. Is OpenAI running some kinda weird AGI or agentic AI social experiment right under our noses? 

Are there covert AIs out in the wild right now, undetectable to most of us? 

We get it. Might sound like we’re wearing a tinfoil hat. 

But like…. y’all. Read that study from the researchers.

Clearly, the most prominent universities and researchers in the world are setting off academic flare guns trying to get our attention and give us all a heads up that advanced AIs masquerading as humans online is gonna be SUPER problematic soon. 

Sheeeeesh slash yikes. 

Alriiiiiiiiight alright. 

We hear ya. 

You show up at Everyday AI cuz we cut through the fluff and give you the facts. 

Like…. We don’t pedal rumors or hype. 

We’re not trying to scare anyone. (So sorry if this is a bit weird.) 

But let’s get real real for a second shorties. 

Like it or not, OpenAI dictates the future of work. So whatever the super advanced agentic/autonomous AI they’re cooking up is gonna be rolled out to the entire business world. 

Think about it. 

Microsoft Copilot? Powered by OpenAI. 

Apple’s new Apple Intelligence? (Partially) Powered by OpenAI. 

So yeah, the companies that control how have tied their future to being powered by OpenAI. And that’s not all. 

Thousands of products, services and essential tools we all legit use every day? (Very likely somehow) Powered by OpenAI. 

👇Look at this incomplete list 👇

↳ JPMorgan Chase: Developed LLM Suite, powered by OpenAI's ChatGPT, to assist nearly 60,000 employees with tasks like drafting emails and compiling reports

↳ Morgan Stanley: Launched AI @ Morgan Stanley Debrief, an OpenAI-powered tool that generates notes and action items for financial advisors

↳ Bain & Company: Partnered with OpenAI to integrate AI into consulting services and expand market reach

↳ The Coca-Cola Company: Engaged with the OpenAI-Bain alliance to leverage AI in operations and consulting services

↳ PwC: Pricewaterhouse Cooper plans to rollout ChatGPT and train over 100,000 of its employees on the technology. 

↳ Salesforce: Tapped OpenAI for a collaboration to enhance Slack with AI-driven features like email drafts and customer-account information

↳ Figure: Collaborating with OpenAI to create specialized AI models for humanoid robots. (They were on 60 Minutes talking about it.) 

↳ BuzzFeed: Partnered with OpenAI to access a large user base for model improvement and enhance user experience with AI-driven features

↳ GitHub: Leveraging OpenAI’s AI codex model into GitHub Copilot, enhancing assisted coding capabilities

↳ Reddit: Partnering with OpenAI to use its models for generating and managing content on the platform

↳ Arizona State University: Has deployed ChatGPT Enterprise to its entire campus. (AI teachers soon, anyone?) 

↳ Moderna: Thousands of employees using ChatGPT enterprise to accelerate medical and pharmaceutical research. 

Why did we list all of those companies out? 

(Not to kill digital trees.) 

Just to give you a small sample of the scope of OpenAI’s influence. Whether you know it or not, so much of your day to day is ultimately going to be influenced by the tech that OpenAI produces. 

And THAT’S why shorties, we gotta sometimes take a lil deep dive on rumors and ramblings on new OpenAI tech, especially when those rumors are being supported/elevated by the CEO and the company itself. 

So if you care about the future of work, you’ve gotta pay attention to writing on the wall. 

Our take? 

We don’t think OpenAI is gonna release GPT-5 anytime soon. They don’t need to. Their newest GPT-4o model is already the best out there. 

But, in the meantime, might we see some new features being rolled out that start to blend human capabilities with our LLM-powered workflows? 

Maybe. 

Maybe not. 

The only thing we know for certain, is we’ll be here every step of the way telling you what it all means. 

Refer Friends to Win a Year of ChatGPT!

(Or your fave LLM like Claude, Gemini, Copilot, etc)