Here’s a preview of what we’ll cover this week: 

Macro: The Fed Hiked and the Long End Listened; The Economy Keeps Giving the Fed Reasons to Stay Firm; The Consumer Forgot to Recede

Markets: Why Anthropic Wants to Slow Down; Defense Tech Is Getting More Expensive; What Happens After A Fed Hike?; Momentum Is Starting to Recover; What Are We Buying?

Lumida Curations: Noam Brown on Conversation Between AI Agents; Brad Gerstner on the $180 Billion Test for AI; David Sacks on the Privacy Gap in AI Chats

This week, I joined Jennifer Sanasie (CoinDesk) at the New York Stock Exchange to discuss the Fed, AI, and crypto.

Here’s what we covered:

Watch the full conversation here.

‘Attention Is All You Need’ is the whitepaper that unlocked the AI cycle. 

Much of what followed has been duct tape. 

I don’t mean that as an insult. Duct tape built half the modern world.

Consider two of the most important techniques driving AI forward today.

First: chain of reasoning.

Instead of asking the model to take one clean shot at an answer, we give it more inference time to decompose the problem, explore alternatives, check its work, and try again.

Step. Check. Revise.

Second: loops.

One model creates. Another evaluates. The first revises. 

The cycle continues until the output survives the critic.

At Meta, Muse is the public example. 

Meta’s AI chief noted that putting an evaluator LLM around a creator LLM, and running the loop repeatedly, produces work comparable to that of a hundred engineers.

Impressive.

Also: It’s more MacGyver duct tape. 

That is what frontier AI increasingly looks like. 

So why do the models keep improving?

Scale.

More data. More parameters. More training compute. More inference compute. Better chips. Better post-training.

But, the breakthroughs have lagged the expectations.

It’s fun to imagine AGI scenarios, from super abundance to dystopia.

But what if AI gets progressively smarter with more data and compute, yet never achieves Einstein or Newton-level insight?

There’s a scenario that is neither ‘super intelligence’ nor dystopian.

AGI becomes the ultimate midtwit. The midtwit to end all midtwits.

We throw more compute at it and get better execution, analysis, automation, and task delegation.

But, it never achieves Steve Jobs or Elon Musk-like creative potential.

In this scenario, no AI invents a fusion reactor or a new theory of quantum gravity. It becomes exceptional at engineering within the world we understand.

No AI religions either.

Computational search is not the same as a conceptual breakthrough.

Newton helped develop calculus and connected falling objects to orbiting planets. He changed how we understand motion.

Einstein asked what he would see if he traveled alongside a light beam. His thought experiments helped reshape our understanding of space and time.

Both changed how we think about the universe. AI can explain why every great idea was obvious, once a human has had it.

Brute-forcing a difficult mathematical problem could be an extraordinary achievement. It would still leave the question of whether the machine had found a solution or invented a new way to think.

Midtwit AGI gets very good at the former. Perhaps it never gets to the latter.

Instead, we get a world of steady incrementalism and productivity growth.

Intelligence becomes a utility on tap. AI does what you want, when you want.

It gives you the conventional wisdom every time. Perfectly. Ask it whether you should start a rocket company and it will recommend a diversified portfolio. The model would have rejected the investment in the company that built it.

An entire consulting firm in your pocket. Mercifully, without the partners.

We spend a lot of time imagining super abundance or an apocalypse. Heaven or hell, with GPUs.

We are under-exploring the Midtwit outcome.

Still a pretty cool world to live in.

In that world, founder agency is paramount. Humans choose what to build, which assumptions to challenge, and which risks are worth taking.

AI makes execution cheaper. Leadership, vision, and judgment still matter.

From a human flourishing perspective, Midtwit AGI might be the best outcome.

Living life becomes a sport.

Accomplishment and meaning still matter. But we have less suffering and more total output.

Not a bad deal.

I recently spoke with a University of Chicago sophomore and Exeter alum who started a poker club backed by prominent trading firms. 

He had already recruited people, attracted sponsors, and built something from scratch.

Then, he asked whether he should start a prediction-market fund or join a long-short hedge fund. Both I explained are terrible ideas.

You want to find the next opportunity while it is still taking shape.

He mentioned that some friends were looking at Y Combinator.

Yes. That’s the move. The entrepreneurial dynamism that is unlocked by AI can’t be overstated.

I am spending more time asking myself ‘How do I build an agent to perform that task?’ rather than ‘Who should perform that task’?

Resume screening, topic ideation, performance review drafting, email scanning.

Learn to find a problem someone will pay you to solve. Convince talented people to join you.

Ship a product and find out whether anyone wants it.

Now is the time to get those AI reps in and learn how to apply it before you wake up in 6 months and feel overwhelmed.

Success in the future will turn on choosing the right problem.

Parenting in the Age of AI

My thinking on teaching entrepreneurship has gone through three stages.

First, give your kids a bit of risk capital (they will probably lose the first round, but at least you can ask for a shareholder update at dinner.). Help them stress their ideas, and give them guidance (choose your friends wisely, work hard, etc.).

Second iteration: Roll-up your sleeves and build a business with them. Don’t preach, do it with them.

Third Iteration: Help them find a peer group of builders. That is what makes a place like Y Combinator arguably more compelling than sending your kids to an Ivy League school.

Learning from a peer group is a different level.

I increasingly think this kind of practical (private) education could be the most valuable part of their development. What’s more — you aren’t actually spending on tuition — you’re getting staked by a VC to build something.

Such is one of the wonders of capitalism - private markets are slowly disrupting public education.

And the idea is not limited to kids. If you are in your thirties and want to build, the opportunity applies to you too.

Keep your personal expenses low. Get close to the people doing the work in SF, Austin, or NYC.

Give yourself room to try.

That is an invaluable education you keep.

Dario wants to slow AI down over catastrophic risks. Sam Altman and Elon Musk agreed. The last time the three agreed was back in 2022.

But, is Dario really slowing down because of “catastrophic risks”? Not at all.

And, today, we have a report in the NY Post that the frontier labs oversold fears to set up a regulatory perimeter.

Here’s why the AI labs want to ‘slow down’.

1. The bills are coming due.

Frontier Labs have enormous compute obligations to fund, and will rely on public capital.

Private investors may fund a story. 

Public investors expect you to deliver against it every quarter.

That means getting into a “beat, beat, raise” rhythm instead of revenue at all costs.

The labs need to turn growing usage into revenue while keeping spending under control. Eventually, that has to produce free cash flow.

And, you don’t produce cash flow by spending on R&D for your next AI model — that is about 30 to 40% of expenses.

You do it by ‘slowing down’. The machines have not become self-aware, but the CEOs and finance departments have.

2. Progress may be slower than advertised.

People I speak with around the frontier labs suggest the pace of development is slower than outsiders think.

The models keep improving.

But much of that improvement comes from more compute, more training, and better engineering.

We have not seen a transformational breakthrough since the Transformer paper.

A call to slow down is a way to reset expectations and shift the goalposts.

3. They want to write the rules before Washington does.

Another reason is midterms. 

The labs want to lead with self-regulation before a new Congress imposes its own version.

The AI Labs are anticipating the Dems take the house (in line with prediction markets).

If regulation is coming, taking the venom out of the sting by talking about regulation is not a bad position to take.

The political incentive is straightforward. Volunteer some restraint now and try to retain more control later.

Will this actually slow compute spending, and slow down spend on semiconductors?

Highly doubtful.

The above explanations change where the spending takes place (mix shift), and it’s a new story.

Every chip that rolls off TSM’s or Micron’s plants gets bought. If Dario, and Altman can’t buy it, Google and Meta or someone else will.

Why? Because the investment is productive. Neo-clouds grow revenue in direct proportion to the number of Nvidia GPUs they can get their hands on.

The Information reports that Shield AI is in talks to raise fresh capital at a valuation of at least $20 billion, following contract wins for Hivemind.

We closed our investment earlier this year at a ~$9 to 10 billion valuation. 

The proposed round would value the company at least twice our entry price. (Past performance may not be indicative of future results.)

We remain excited about defense tech.

In private markets, 2018 to 2021 was the crypto cycle. 2023 to 2025 was the AI cycle. The Defense Tech cycle is underway now. We think investing in private unicorn AI businesses with zero revenue is a terrible idea - esp with the Anthropic IPO.

We are now evaluating one of the leading drone companies backed by the same VC that invested in Anduril and SpaceX early. If you’d like more details on this deal, you can sign up here. (This isn’t an offer to buy or sell securities)

If you are an accredited investor, or qualified purchaser, you can get on our deals communication list through Lumida Deals. The windows on these private deals can be short, so joining beforehand matters.

(Disclosure: This communication is provided for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to buy any security or interest in any fund or investment. Any such offer will be made only to eligible investors and only by means of definitive offering documents, which should be reviewed carefully in their entirety, and only in jurisdictions where permitted by law. Investments in private, early-stage companies are speculative and involve a high degree of risk, including the risk of illiquidity and the total loss of capital; such investments are suitable only for qualified investors who can bear these risks. Any statements regarding the company's business, strategy, or prospects are forward-looking, are based on information believed to be reliable but not independently verified, and are subject to change without notice. Past performance is not indicative of future results, and there is no guarantee that any investment objective will be achieved. Nothing herein should be construed as investment, legal, or tax advice; recipients should consult their own advisors before making any investment decision.)

Separately, thank you to everyone testing our Tax Mitigation App.

For readers unfamiliar with it, “Second opinion” uses AI to identify potential tax mitigation opportunities for your unique situations. The service is still in beta and currently free.

Try it here and share your feedback. We would especially like to know what it missed.

Lumida offers strategies designed to mitigate capital gains, estate, and income taxes. Email [email protected] if you would like to discuss your situation.

Use of the app does not constitute tax advice. Review potential strategies with your CPA or Lumida advisor.

On Wednesday, the Fed raised rates 25bps to 3.75%–4.00%.

The vote was unanimous, 12–0.

Warsh pointed to three reasons for the decision: stronger growth, insufficient progress on inflation, and greater geopolitical risks.

The Fed’s updated projections pushed a full return to 2% inflation out to 2029. 

Markets had largely priced in the rate hike before it happened, so there wasn’t much reaction.

A hike strengthens the Fed’s credibility on inflation. And that can help bring the long end down. 

And, that’s what happened. The 10-year Treasury yield fell back below 5%.

We think the 10Y has topped, and it should drop further as oil prices start easing.

Does a rate hike mean we have a new hiking cycle?

No. Just like the Fed introduced a salvo of ‘adjustment cuts’ we are seeing an adjustment hike.

And, a small rate hike will not reduce the earnings power of companies that are growing that don’t need debt to grow.

The bigger issue will be smaller companies with heavy debt loads that have significant refinancing needs.

Last year, labor market fears helped Powell avoid any hikes. Warsh didn’t have the same cushion.

The labor market remains healthy.

Initial jobless claims fell to 196,000 in the week of September 11. Claims have now come in below 200,000 five times this year, versus once in 2025.

The four-week average of continuing claims also fell for a fourth consecutive week, reaching its lowest level since January 2024.

Few layoffs, and fewer people staying on unemployment benefits. That is a supportive backdrop for household income.

Labor market is intact thanks to the broad strength across the economy.

Atlanta Fed GDPNow raised its Q3 real growth estimate from 4.4% to 5.1%.

The revision was driven by real consumer spending growth, which was revised from 3.6% to 4.1%.

The Death of the US Consumer remains greatly exaggerated.

Business investment remains strong too.

September retail sales number came in strong as well, with same-store sales rising 8.4% YoY in the week of September 11, versus a 5.8% average in 2025.

Bank of America shared their consumer pulse report this week, and it had valuable insights on the US consumer. 

August household card spending rose 4.5% YoY, and 0.9% MoM. Excluding gasoline, growth was still 3.7%.

More interesting is who is spending.

Lower-income households’ discretionary spending grew 5.7%, nearly matching 5.9% for higher-income households. Middle-income spending rose 5.1%.

Both ends of the K-Shaped economy are spending.

And the debt-funded spending argument gets weaker. 

More households are paying their credit cards in full.

Among those carrying balances, utilization declined year over year and sits near pre-pandemic levels.

Lower-income households also saw wage growth for a second consecutive month, adding to their purchasing strength.

Real median household income reached a record $87,463 in 2025.

Those are annual figures, but they show households entered this period with a stronger income base.

Earnings should keep flowing.

The S&P 500 ended the week flat, while small caps lagged, down 1.5%. 

This was the first Fed hike week after a series of cuts in the last two years.

Here’s a study of the six tightening cycles since 1994. 

The S&P 500 has declined an average of 3.2% in the month following the hike. It rose in just 17% of those instances. 

Financials and health care were the most vulnerable sectors, down 6.4% and 5.6% in the first month, on average, respectively. 

Note: The Fed, in our view, is executing ‘adjustment hikes’ — they are not embarking on a full blown tightening cycle - so take these stats with a grain of salt.

If / when energy prices come down, there’s less reason for the Fed to hike.

Three months out, the average return was still negative at 2.3%.

By six months, it improved to a 6.4% gain, still trailing averages.

Does that mean you should expect turbulence?

AAII bearish sentiment rose to 53.3%, its highest reading since the 2025 tariff tantrum. 

Bullish sentiment fell to 28.8%, its lowest since last September.

A few weeks ago, we were concerned about crowded optimism. It's the opposite today.

You sell in Euphoria, you buy in fear.

Momentum names are showing early signs of recovery after hitting deep oversold levels in early September. 

Twelve-month momentum factor is bouncing from oversold levels.

You can also notice the same strength in the 3-month momentum chart. 

This is an area where we want exposure. 

Momentum factor contains names in semis, and industrials, which, after months of euphoria, have now returned to reasonable valuations, and in some cases, reflect fear.

For the first half of the year we spent time noting that Industrials were way over-priced.

They’ve come down quite a bit now. Still prefer semis over industrials here, but one can make the case for Industrials now.

We remain bullish on semis.

Competition, sovereign investment, and broader adoption give demand several sources of support. 

And, it’s not only us. 

Is Momentum Back?

You can analyze factor performance through Lumida Strategies on the Lumida Invest app.

Here are the top 5 performing strategies for the week. Notice how it is all momentum. 

You can also explore what each strategy has bought and sold to adjust your exposures. 

Our goal is to deploy AI agents that manage your portfolio based on these types of strategies by year-end. 

We bought Medical Devices (IHI is the ETF). This gives us value, quality compounders, and exposure to the aging demographics theme.

After an approximately 11% decline over the past month, it is finding a base, and is available at all-time low valuations.

You can now view Lumida Curations at the Lumida Invest App. Curated insights delivered on time. Download the app here. 

OpenAI’s Noam Brown argues that rigid multi-agent systems isolate agents, duplicate work, and force them to guess when instructions are unclear.

View Curation

Brad Gerstner argues that the top three AI labs must grow their combined revenue run rate from roughly $100 billion to at least $180 billion by year-end to sustain the AI trade.

View Curation

David Sacks argues that AI conversations deserve at least the same legal protections as email, especially as users share sensitive medical, legal, and personal information.

View Curation

Not Subscribed Yet? Don’t miss out on future insights—subscribe to the newsletter now!

For real-time updates, follow us on: 

X | Telegram | Youtube | TikTok | News | Ram’s X | Lumida Health | Lumida Tax | Lumida Memes

As Featured In

Disclaimer: Lumida Wealth Management LLC (‘Lumida”) is located in New York, NY, and is an SEC registered investment adviser. SEC registration does not constitute an endorsement of the firm by the Commission nor does it indicate that the adviser has attained a particular level of skill or ability. Lumida only transact business in those states in which it is registered, or qualifies for an exemption or exclusion from registration requirements. Any direct communication by Lumida with a prospective client will be conducted by a representative that is either registered or qualifies for an exemption or exclusion from registration in the state where the prospective client resides.

The information in this material has been obtained from sources believed to be reliable. While all reasonable care has been taken to ensure that the facts stated in this material are accurate and that the forecasts, opinions and expectations contained herein are fair and reasonable, Lumida, Inc. and Lumida Wealth Management LLC (collectively Lumida) make no representations or warranties whatsoever the completeness or accuracy of the material provided, except with respect to any disclosures relative to Lumida. Accordingly, no reliance should be placed on the accuracy, fairness or completeness of the information contained in this material. Any data discrepancies in this material could be the result of different calculations and/or adjustments. Lumida accepts no liability whatsoever for any loss arising from any use of this material or its contents, and neither Lumida nor any of its respective directors, officers or employees, shall be in any way responsible for the contents hereof, apart from the liabilities and responsibilities that may be imposed on them by the relevant regulatory authority in the jurisdiction in question, or the regulatory regime thereunder. Opinions,forecasts or projections contained in this material represent Lumida’s current opinions or judgment as of the day of the material only and are therefore subject to change without notice. Periodic updates may be provided on companies/industries based on company-specific developments or announcements, market conditions or any other publicly available information. There can be no assurance that future results or events will be consistent with any such opinions, forecasts or projections, which represent only one possible outcome. Furthermore, such opinions, forecasts or projections are subject to certain risks, uncertainties and assumptions that have not been verified, and future actual results or events could differ materially. The value of, or income from, any investments referred to in this material may fluctuate and/or be affected by changes in exchange rates. All pricing is indicative as of the close of market for the securities discussed, unless otherwise stated. Past performance is not indicative of future results. Accordingly, investors may receive back less than originally invested. This material is not intended as an offer or solicitation for the purchase or sale of any financial instrument. The opinions and recommendations herein do not take into account individual client circumstances, objectives, or needs and are not intended as recommendations of particular securities, financial instruments or strategies to particular clients. The recipients of this material must make their own independent decisions regarding any securities or financial instruments mentioned herein and should seek advice from such independent financial, legal, tax or other adviser as they deem necessary. Lumida may trade as a principal on the basis of its views and research, and it may also engage in transactions for its own account or for its clients’ accounts in a manner inconsistent with the views taken in this material, and Lumida is under no obligation to ensure that such other communication is brought to the attention of any recipient of this material. Others within Lumida may take views that are inconsistent with those taken in this material. Employees of Lumida not involved in the preparation of this material may have investments in the financial instruments or securities (or derivatives of such financial instruments or securities) mentioned in this material and may trade them in ways different from those discussed in this material. This material is not an advertisement for or marketing of any issuer, its products or services, or its securities in any jurisdiction.