So, Anthropic is preparing to go public.

Its backers told the Financial Times they expect an October listing at a valuation of two trillion dollars or more. That's ahead of SpaceX, in just a few months.

At this stage we might as well use the numbers my 5-year-old uses.

Is sextillion bigger than godzillion? No idea… both sound great.

I do not know what anything should cost. And as we established in Fomonomics, neither does anyone else.

I didn't want to dwell on the ridiculousness of the figures, but I must, just for a bit. In May, Anthropic's valuation was $965 billion. Now, 2 trillion. A trillion dollars of value created between two board meetings.

Good stuff!

For reference, that is roughly the GDP of Switzerland

Fomonomics on full display: a record offering, institutions queuing with other people's money, and, somewhere in the queue, your pension fund, waving.

But this letter is not about the craziness of the modern economic model. We already covered that.

This is a case study, and a beautiful one, because inside the small terrarium of the AI race we get to watch a controlled experiment on the oldest question in management: does focus actually work?

OpenAI built the most famous product on the planet and tried to do everything for everyone: it tried to make you a therapist, a girlfriend, make your homework, and your imagination.

Anthropic did nearly the opposite and tried to stay focused. Mainly, it sold tokens to businesses through APIs and cloud contracts.

Claude has roughly five percent of ChatGPT's consumer audience. Eight of the Fortune 10 are customers. Revenue is up from $787 million in a quarter to $11.5 billion in the same quarter a year later, and the run rate passed $65 billion in July. That is what a focused strategy looks like.

The greats of our craft put focus above almost everything.

Michael Porter, more or less the first strategy celebrity, in 1996: "The essence of strategy is choosing what not to do."

Steve Jobs, the first "cool CEO," cutting Apple's product line by seventy percent in 1997 to save the company: focus means saying no to the hundred other good ideas, and innovation is saying no to a thousand things.

Our patron saint, Peter Drucker, got there in 1964: concentration is the key to economic results. He also added that no principle of effectiveness gets violated more often.

He wrote that sixty years before your company announced its twelve priorities.

As we said in Lesson 1: twelve priorities is not a strategy; it is a wish with a fancy font.

So the evidence is overwhelming, the advice is sixty years old, and the AI terrarium just ran the experiment live with a two-trillion-dollar result.

Because focus has a psychological price that is paid in the meeting room.

Focus means naming the things we are not going to do. Out loud. With the owners of those things sitting at the table, maintaining eye contact, remembering.

Every real strategy kills somebody's project, somebody's budget, somebody's little kingdom. A list of twelve priorities kills nothing and upsets nobody.

It has always been this way, but in the last few years it got worse. AI has made ideas easy. Anyone in your organisation can generate ten concepts and three working POCs in a couple of days and send them in an email that sounds "strategic."

The cost of producing a plausible initiative has collapsed to zero, so the volume has exploded, and your brain as a manager is getting DDoSed with ideas that all demo well.

All of them. Every single demo works now.

It is a miracle.

Meanwhile, the receiving hardware, your brain, is degrading.

Gloria Mark, professor of informatics at UC Irvine, has measured attention on screens since 2004, first with stopwatches, then with logging software. In 2004, people averaged two and a half minutes on a screen before switching. By 2012, 75 seconds. Today, 47 seconds, replicated by five independent studies.

Recovering full focus after one interruption takes about 23 minutes.

Sit with that arithmetic.

Focus lasts under a minute; repairs take twenty-three. We are running civilisation on a machine that breaks forty times faster than it fixes.

And the worst finding: we interrupt ourselves about as often as anyone interrupts us.

There is no villain to fire.

The call is coming from inside the skull.

Add the Stanford result that heavy multitaskers get worse at filtering, not better (practice makes worse, a first for our species), and the University of Texas study showing that a phone lying face down on the desk drains cognitive capacity even when switched off, meaning the phone doesn't need to be on to win; it just needs to exist, and the picture is complete.

So focus is now a two-front war.

Strategic focus: fewer bets, making sure every choice is evaluated and made.

Attentional focus: the capacity to actually work on them.

You will notice the modern corporation loses both fronts simultaneously.

It will help you to at least work on your own attentional focus before you attempt to focus an entire organisation.

When you are making decisions, drafting strategy, whatever it is, list what it explicitly declines.

No noes, no strategy – you are literally not allowed to call it that.

I am not kidding.

Literally on every slide: on objectives, which objectives we declined; on KPIs, which KPIs we declined; on initiatives, which initiatives we declined.

This is one of the best training techniques for your brain to stay focused.

It is also an excellent way to find out, in real time, which of your colleagues have never declined anything in their lives.

This one works for me quite well.

I killed all notifications on my phone.

WhatsApp, email, apps, everything.

Only calls remain, and even those are on a very light buzz.

Nobody has died.

Several people believe I have become a better person; in reality, I have merely become unreachable, which in modern corporate life is indistinguishable.

Kushlev and Dunn found people limited to checking email three times a day reported significantly lower stress.

A later study found the same for batched phone notifications.

Gollwitzer, a professor of psychology at New York University, developed implementation intentions, one of the most replicated effects in applied psychology: goals with a specified trigger ("when the laptop opens, forty minutes on Word before email") get achieved at far higher rates than goals stated as intentions.

An intention is a hope.

A when-then is an appointment.

Your brain, it turns out, respects appointments and ignores hopes, exactly like everyone else in your calendar.

Train your brain to stay focused, then train it to make focused decisions, then have a focused plan.

In that order.

Otherwise it doesn't work, and you end up with the standard alternative: a distracted brain, approving unfocused decisions, inside a plan with twelve priorities, wondering why the focused competitor is worth two trillion.

Talk soon,
Sultanbek