Why it exists
The information was there years before the price was
In July 2010 a bitcoin cost less than ten US cents. In June 2018 OpenAI published a language model and almost nobody outside the field noticed.
Neither was a secret. Both were being argued about in the open by small groups of serious people. Most of us were reading something else.
Next Wave is built to find that gap: the distance between how big something will become and how little has been built or priced so far. Sometimes that means a fringe idea. Sometimes it means a trend everyone knows, followed two or three steps past the headline. Example: If AI needs energy to scale (obvious), what runs short next, and what follows from that (less obvious)?
What you get
Three to five of these, every second Tuesday
This one is from the first edition, shortened to fit.
Fringe
1. Tools for editing mitochondria
Magnitude ●●●●●Asymmetry ●●●●●Evidence ●●●●●Mainstream 28/100
On 9 September 2026 the UK’s ARIA agency committed up to £66 million to 19 teams. They will try to insert genetic instructions into mitochondria, the parts of cells that make energy, and keep those instructions working. Nobody has done that reliably yet.
Why now. Nineteen funded teams are now running the same kind of experiment at once. That creates a chance to sell shared tools and compare results, years before any treatment exists.
The secret
The funded teams need reliable ways to measure whether genetic instructions arrive and keep working. Shared tools and comparable tests could sell long before a treatment does.
Ways to play
To build
A service that tests each team’s delivery method the same way, so results can be compared. The 19 labs are the first customers.
Picks and shovels
Quality checks and specialist analysis sold to researchers.
The sceptic
Nineteen grant recipients are not nineteen paying customers. The programme could fund its own tests and comparisons. Editing genes in one organism proves no transferable method for another.
45%likely
Prediction · by 20 Sep 2028
At least one funded team publicly reports delivering working genetic material inside mitochondria, and an independent laboratory reproduces it.
How it is made
An AI-written letter
This is not a newsletter curated by a person. Every edition is researched and written by AI, attacked by a second AI whose only job is to challenge and reject, then assemble and send the output in plain English.
I built the system. I do not write, choose or approve what it sends, and I do not read an edition before you do. It can be wrong. Every idea links to its sources so you can check it, and that is what you should do before relying on any of it. If you find an error, reply to the letter. Replies go to me, not to the AI: it never reads them and never answers them. Errors I confirm are corrected in a later edition.
It is general information, not financial, legal or investment advice. It recommends nothing and knows nothing about your circumstances. See the privacy notice and terms.
There is also no track record. The first edition went out in September 2026 and nothing here has been proved right. That is why every idea carries a dated prediction with a probability: in two years you will be able to check whether this was worth reading.