😺We took an MIT AI course!

PLUS: GPT-4.5 passed the Turing Test?!

Welcome, humans.

What do you know—ChatGPT just made image editing available to all free users. Cue the obligatory marketing tweets about Sam’s GPUs: “We’re melting, we’re melting…”

Actually, this person had a pretty good take on where all those magic new OpenAI users were coming from: since the free version has a limit of 3 images per day… why not make more accounts?

Anyway, did anyone in your life use OA’s new imgen to totally prank the pants off you and your home owner’s insurance provider, like this Redditor? Wish we saw this yesterday, if so we would have flagged it for all you wanna-be pranksters…

Here’s a few other interesting imgen uses:

Sharpening a blurry photo…

Bring your childhood drawings to life (more drawings)…

Comic-ify your daily struggles…

Or our personal favorite: improving your headshots! For example, check out Mason from our team at Technology Advice testing the tool to improve his profile pic:

It’s a me, Mario Tyson!

Even some graphic designers are embracing the new tools to work smarter, not harder. After all, what’s so different about using ChatGPT to aid your design versus Canva? In the immortal words of Tim Gunn: Designers, make it work…

Here’s what you need to know about AI today:

We recap an awesome AI MIT course we took.
ChatGPT hit 500M weekly users and 20M paid subscribers.
New tech enabled a paralyzed woman to talk at 90+ words per minute.
GPT-4.5 passed the Turing test.

MIT xPRO's new Generative AI course will make you an AI designer (even if you can't code)

Ever feel like everyone's talking about AI, but nobody's teaching you how to actually use it?

Have no fear, because MIT xPRO's “Driving Innovation with Generative AI” is here. We got the opportunity to take part in the first cohort, and we learned A TON.

Yes, even those of us who eat, sleep, and breathe AI every day can learn something from this course.

What you'll ACTUALLY learn: This isn't your typical “watch videos, take quizzes” online course. Over six weeks, you'll dive into the full spectrum of generative AI technologies guided by professors from MIT’s world-renowned faculty.

Here’s what’s included:

The 6-week course teaches you to design AI solutions without coding.
It covers everything from language models to image generation to drug discovery and everything in between.
You learn from 12 different MIT professors through practical assignments.
And you'll build real AI solutions that solve problems in your specific industry.

For example, we share two of the six projects we built during the class in our full write-up on the website. Here’s one of them—a customGPT called Decider! Try it out and let us know what you think.

Got decision fatigue? Same. That’s why we built this to help rank our priorities, rank our choices, and then let the AI decide the next right move to hit our goals.

Here’s the week by week breakdown:

AI Landscape Fundamentals: Neural networks, data requirements, and how generative AI differs from traditional machine learning.
Visual Generation Models: The mechanics behind image generation, understanding latent variables, and analyzing inputs/outputs of systems like DALL-E and Midjourney.
Large Language Models: The evolution of language models, their current limitations, and future applications across various fields.
Specialized Applications: How genAI transforms coding, chemistry, and 3D design through molecular property prediction and structural design.
Human-AI Collaboration: Techniques for creative expression and designing AI that complements human capabilities.
Pro-Human AI Development: Addressing biases, implementing feedback-driven reinforcement learning, and aligning AI with user goals.

Our main takeaway? That building AI isn't just for engineers anymore. The course focuses on conceptual design rather than technical implementation, making it accessible for product managers, UX designers, and professionals from any industry looking to leverage AI.

Now, we didn’t go into the course with any set projects in mind, but we definitely recommend you do—if you have an AI project you’ve always wanted to build, this is a great jumping off point to learn how to design it.

Or you can wing it, like we did—you never know what kind of creative stuff you’ll come up with. You can read more about our experience taking the course here.

Final verdict–who should take this course? If your company is struggling to move beyond “we should be using AI somehow,” this could be your chance to become the in-house expert. The 3 Continuing Education Units (CEUs) from MIT won't hurt your resume, either!

For busy professionals wondering if it's worth the time investment, the course structure lets you access all content online and watch videos on the go. Perfect example: We were able to finish the course while writing this newsletter full-time!

It'll set you back $2,979, but hey—that's probably less than your company spends on coffee each month (so just get your boss to pay for it!).

You can learn more or enroll on MIT xPRO website—and here’s a 15% coupon that’s valid until April 11 from us!

Prompt Tip of the Day

These 7 prompting hacks transform mediocre AI responses into “yeah, I can work with that.” We've been secretly using them to level up our own AI game:

Be ultra-specific: “Write a 3-paragraph email to clients about our AI workshop on June 15th” beats “Write an email” every time.
Assign personas (this is hit or miss, but helps with technical topics): “Respond as a senior financial analyst” instantly upgrades financial advice.
Talk like a human: Conversational language > robotic mega-prompts.
Add context: “For a startup targeting Gen Z in Miami” changes everything.
Request exact formats: “Create a 5-point checklist” gets you... an actual checklist.
Stay polite but direct: “Please” helps, but paragraphs of politeness don't.
Iterate, iterate, iterate: This is the secret sauce! Start simple, then say “Now make it more persuasive” or “Rewrite for beginners.”

Our favorite of these seven? The last one. We never stop at the first response when working with AI. Treat your AI like clay—molding it with follow-up prompts until it's exactly what you need.

FROM OUR PARTNERS

💼 Want to become an AI Consultant?

This stat is WILD: The AI consulting market is projected to explode from $6.9 billion to $54.7 billion by 2032. Want to capture this opportunity?

So how do you, an AI enthusiast, become a handsomely compensated AI consultant?

Well, The AI Consultancy Project from Innovating with AI has already helped 300+ students build successful consulting businesses.

Current students love:

Tools and frameworks to find and serve clients.
6-month roadmap to building a 6-figure business.
Methods to land your first client in as little as 3 days.

As a Neuron reader, you get early access to the next enrollment cycle!

Take advantage of the AI boom—request early access to The AI Consultancy Project ASAP!

Treats To Try.

OpenAI Academy is OpenAI’s free educational resource for helping you learn how to use ChatGPT and other AI tools better.
Apple Intelligence is now available on the Vision Pro letting you polish your writing, create custom images, and find photos.
Luma Labs’s new Camera Motion Concepts lets you create and combine custom camera movements that work across different styles—try here.
Higgsfield gives you access to over 30 professional camera motions like Dolly Zoom, Dutch Angle, and FPV Drone to enhance your videos.
GeometryCrafter extracts precise 3D geometry from your videos, converting regular footage into depth maps and point clouds with one click (free to try).
Triplane Turbo instantly transforms your text prompts into detailed 3D models with textures in just over a second—try this demo.
Keyword Research is a simple keyword research tool with search volumes, difficulty scores, and 12-month trends without requiring signups or payments (European formatting tho, FYI).
GemmaCoder3-12B is an open AI tool that improves your coding by solving complex programming challenges with 50% better performance (read more)—here’s how to use it with LM Studio.

See our top 51 AI Tools for Business here!

Around the Horn.

ChatGPT says it hit 500M weekly users and hit 20M paid subscribers, up from 15.5M at the end of 2024, which The Information estimates is worth about $415M / mo and $5B / year (based only on $20 / mo tier, so likely higher)—but their goal is to hit $12B+ in 2025.
In addition to ChatGPT’s traffic boost, it looks like Gemini, Claude, and DeepSeek all got a boost, too—Gemini’s up to 10.9M average daily visit, and Claude’s up to 3.3M, while DeepSeek is tied with Grok at 16.5M visits.
Agility Robotics wants to raise $400M at a valuation of around $1.75B—as a reminder, they make this little guy.
Meta plans to launch new AI / AR galsses for $1K called Hypernova, which will be able to run apps and display photos with hand gestures.
A new study may have just proved OpenAI used unlicensed books from O’Reilly Media as data to train ChatGPT-4o due to a strong recognition score (82%) of the non-public content, using a method called DE-COP to identify the content.
A new Brain-to-speech technology enabled a completely paralyzed woman to communicate at near-conversational speeds, turning neural signals directly into her natural-sounding voice at 90+ words per minute with only a one-second delay.
GPT-4.5 has now passed Alan Turing's original 70% benchmark for machine intelligence (known as the Turing test)—it was judged as human 73% of the time, outperforming actual humans.
Here are the top AI papers from last week (courtesy of Elvis Saravia and DAIR, who offer AI courses for $39 a month).

FROM OUR PARTNERS

Beta Acid turns your wild AI ideas into actual products

Founded by a long-time startup CTO, Beta Acid’s team designs and builds MVPs with lean efficiency and zero corporate nonsense. Your brilliant idea + their technical wizardry = market-ready innovation. Serious skills, fun approach. Go get sh%$ done!

What can you build together?