FORMA's Five Things.

FORMA Programs

You have learned to talk to AI.
Now learn to build with it.

The next Build With AI session is forming now. Tell us what you want to get your hands on and we will build the session around it.

The World Cup Just Became
AI's Biggest Stage

For the first time in World Cup history, every team in the tournament has access to the same AI. All 48 of them.

FIFA and Lenovo built a tool called Football AI Pro. It analyzes hundreds of millions of data points and delivers tactical intelligence to any coaching staff that wants it. That matters because until now, the gap between a well-funded European program and a smaller nation was partly a data gap. Germany had analysts. Cape Verde did not. That is changing.

But it goes deeper than tactics. The ball itself has a sensor firing 500 times per second. Players are scanned into 3D avatars so offside calls can be made with precision no human linesman could match. Stadium digital twins let security teams monitor crowd movement across 16 venues in three countries in real time.

The World Cup has always been where technology gets its biggest test. Color television. GPS tracking. VAR. Each debuted here before it went everywhere else. AI is next. And this time, the underdog gets a seat at the table too.

Read more →

Publishing Just Entered Its First
True AI Era. The Numbers Are Stark.

Two economists tracked every new book release on Amazon between 2022 and 2025. What they found should make anyone in a creative industry pay attention.

New releases tripled. Monthly volume went from roughly 100,000 titles to over 300,000. In some categories it grew by a factor of ten. Average quality dropped. Books with detected AI content get fewer ratings, lower star scores, and weaker sales than human-authored titles. By 2025, more than half of all new releases showed AI involvement.

Here is what did not happen: established authors were not displaced. Writers who were publishing before the LLM era are still publishing, and their work still outperforms the flood. The quality decline is coming from new entrants who arrived after the tools did, not from writers using AI to work faster.

The story is not that AI ruined books. Every medium that got easier to enter went through this. Blogging. YouTube. Podcasts. Publishing is just the latest.

Read more →

The Company That Made AI Images
Just Built a Medical Scanner

Nobody saw this coming. Midjourney, best known for turning text prompts into art, announced last week that it is building a full-body medical imaging device.

The device is called the Midjourney Scanner. You step onto a platform, descend into a shallow pool of water, and pass through a ring of half a million ultrasonic sensors. Sound waves fire through your body from every angle. A compute cluster reconstructs the data into 3D images of your muscle, fat, bone, and organs. The whole scan takes about 60 seconds. No radiation. No magnetic fields.

The device has no FDA clearance yet and the first location is still a couple of years out. But Midjourney has no outside investors, is profitable, and has always moved deliberately. The same instinct behind their image tools, using computation to see what the eye cannot, is exactly what they are applying here. Whether it works at scale is an open question. That they built a working prototype is not.

Read more →

The Fable Freeze Is Still On.
And It Is Bigger Than One Company.

Last issue we covered the moment the US government ordered Anthropic to pull its two most powerful models offline. Ten days later, Fable 5 and Mythos 5 are still dark.

Bruce Schneier is a cybersecurity expert, a fellow at Harvard's Kennedy School, and one of the most widely read voices on technology and security in the world. Writing in The Guardian this week, he argues that the government's move will not solve anything. The problem is not Fable. The problem is that AI capabilities are climbing every year, and no single ban changes that trajectory.

What the Fable freeze actually revealed is a governance gap. There is no global framework, no coordinated authority, and no agreed process for deciding when an AI model is too powerful to release. The US used export control law, written for weapons and foreign adversaries, to pull a subscription service used by hundreds of millions of people. With almost no notice. On a Friday afternoon.

Schneier calls it what it is: the opening of a Pandora's box that no one has a plan to close. Other frontier models are months behind Fable. Open source versions are less than a year behind. Whatever comes next, this moment is a preview of a conversation that governments, companies, and the public have not had yet but will need to.

Read more →

You Are Asking AI About Your Business.
It Has Never Seen Your Business.
What Is RAG, and What Will It Solve?

The AI you use was trained on the internet. Books, articles, research, billions of words from everywhere. But not your files. Not your processes, your client history, or your institutional knowledge. When you ask it about your business, it is guessing. This is the problem RAG was built to solve.

RAG stands for Retrieval-Augmented Generation. Instead of relying only on what a model learned during training, RAG connects the AI to a specific set of documents, a knowledge base, or a file system. When you ask a question, the system retrieves the relevant pieces from that source first, then generates an answer based on what it actually found. The AI stops guessing and starts reading.

Think of it this way. A standard AI is like a brilliant consultant who read everything but has never been to your office. RAG is what happens when you hand that consultant your actual files before the meeting. Same intelligence. Much better answer.

The difference between an AI that knows your industry and one that knows your company is enormous. RAG is how you close that gap. You do not need to retrain a model or build anything from scratch. You just need to connect the right documents to the right questions. Most people using AI at work have the brilliant consultant. They just have not handed over the files yet.

FORMA Tools

Not sure which AI model is actually right for your work?

Six questions. A personalized model recommendation. No benchmark tables, no guesswork. Just a clear answer based on how you actually work.