PhD CPA Assistant Professor at @TCU. Argues with people who agree with me. Official Matt. All views are my own. (He/Him) @ menzer on the other one
Joined July 2010
- Tweets24.7K
- Following850
- Followers1.1K
- Likes12.5K
Pinned Tweet
Just putting this here for future reference anytime someone says: "but I have never been audited on it"
#taxtwitter
AICPA SSTS 2.1.8(a)
Tyler Menzer retweeted
This is one of the most important pieces of social science on human and AI interaction that I’ve seen in some time. Experts follow agentic advice the vast majority of the time (causally), but only a tiny fraction credit the agent—majority think they came to decision on their own volition.
Reminds me of the classic Nisbett and Wilson study on people following outside influence without realizing it
“ But as more work is handed to AI—including AI safety research—these judgment calls are the main levers for people's own values to influence outcomes.
If people defer to AI without noticing, then "human in the loop" might provide less assurance than people expect.”
What if instead of direct personalized pricing it's hidden behind customer service. The price is high for everyone all the time, but when you use your agent to complain certain people are treated much more favorably than others, lowering their effective price. Pricing through CSR
wrote about AI AS THE INFINITE WORK MACHINE, creating a world where we're all just constantly deploying agents to fight against everyone else's agents, running constantly to end up in the same place:
bloomberg.com/news/newslette…
Tyler Menzer retweeted
In a recent NYT debate, Paul Kedrosky claimed that 80% of Anthropic revenues come from 8 companies. He claimed Meta alone makes up 10%.
As Azeem Azhar pointed out, this concentration was declining over time and we'd have an answer in the S-1.
The S-1 tells us:
"No single customer represented more than 10% of our revenue for the nine months ended September 30, 2026. Approximately 79% of our revenue came from customers outside of our top five customers for the three months ended September 30, 2026, up from approximately 64% for the three months ended March 31, 2026."
So that resolves it. Concentration is below the levels Kedrosky warned of, and falling.
nytimes.com/2026/10/16/opini…
sec.gov/Archives/edgar/data/…
I have been saying this for awhile. Not just customer service. Think AI ads, personalized marketing tools, corporate strategy and pricing. Many things businesses do is competitive. A lot of tokens can be spent for a costly arms race that leads to no profit
As @tracyalloway writes in today's Odd Lots newsletter... The logical endpoint is my agent battles some company’s algorithm until they reach the same equilibrium we would reached before AI, except we both spent a lot on tokens and the whole thing is value destroying.
I'll say, because I was a bit glib with my first Anthropic S-1 tweet, the f/s look better than I would have expected. Positive operating cash flow and a strong b/s. The one thing I don't really like is that it feels like looking at half a f/s...
1/
Anthropic currently has just $0.5B of datacenter fixed assets on its books and another $0.5B in progress.
For being 1 of the big 2 in an industry that is projecting to spend $1T in capital just this year, that is tiny.
2/
"gross margins are above 80%"
Non-gapp numbers tell a story, but GAAP comes for us all eventually
I feel like class actions should really be seen as classical pigouvian taxes, where the societal benefit can come primarily from imposing the cost, regardless of what is done with the money.
🚨🎮 UBISOFT'S $2 MILLION SETTLEMENT. GAMERS CLAIMED JUST $38,633!
Gamers, this is interesting. Very interesting. 🤔
Remember The Crew? The racing game Ubisoft shut down in 2024, leaving people who purchased it unable to play?
Um, that decision led to a class-action lawsuit, and now we're getting a closer look at the numbers.
According to an October 16 court filing, Ubisoft established a $2 million settlement fund for affected consumers.
Now things get interesting.
Approximately 1.8 MILLION people were included in the settlement class. Yet only 2,951 valid claims were submitted, totaling $38,633 in cash and Ubisoft store credit.
Meanwhile, the attorneys are requesting $800,000 in fees, plus litigation expenses.
Before you start sharpening your pitchforks, the attorney fees would be paid separately by Ubisoft, not deducted from the $2 million consumer fund. And the court hasn't approved them yet.
But seriously..
$2 MILLION AVAILABLE. $38,633 CLAIMED.
That's less than 2% of the available funds.
This part bothers me.
The lawsuit wasn't simply about getting a few bucks back. It challenged what consumers were led to believe they were purchasing in the first place.
Ubisoft shut down the servers. The game became unplayable. And the people who purchased it were left with a product they could no longer use.
The proposed settlement offers qualifying claimants $7 in cash or $15 in Ubisoft store credit.
One question.
HOW DOES THAT BRING THE GAME BACK?
It doesn't.
And that's the bigger problem with digital ownership. We keep having the same discussions about compensation AFTER access disappears, rather than protecting consumers BEFORE it happens.
To be fair, Ubisoft denies wrongdoing, and the settlement still requires final court approval. The hearing is scheduled for November 13.
All this deserves a much larger conversation.
When you purchase a game, should a publisher have the ability to make that game permanently unplayable simply because maintaining its servers is no longer part of the business plan?
And if that happens, is a few dollars really a fair remedy?
I'm curious what you think. This isn't only about Ubisoft anymore. It's about the future of the games we're all paying for.
Official settlement information and court documents:
crewgamesettlement.com/
🤓 Never underestimate the power of nerds.
If you like my content, follow me. Game On.
#Ubisoft #TheCrew #DigitalOwnership #ConsumerRights #GamePreservation #GamingNews #EthicaGaming
Anyone saying billionaires shouldn't exist should be forced to use Oprah, Beyonce, and Taylor Swift as their first examples.
If it is really true that no one can legitimately make a billion, then surely they are just as exploitative?
How would people feel if the CBO scored (unchallenged) Trumps tariffs as raising no revenue under the theory that they were unconstitutional?
Or scoring a Dem repeal of the same tariffs as costing $0 because they are illegal
Tyler Menzer retweeted
Interesting setup! In my opinion, questions of diversity, gameability, and algorithmic monoculture remain, even if AI reviewing is getting better. Our ICML paper might be of interest: xitter.cf/joabaum/status/2050278…
Can you boost your AI review scores by asking an LLM to rewrite your paper?
Yes! We call it paper laundering
Our @icmlconf spotlight paper argues current AI reviewers aren't ready to automate peer review, and outlines what a science of peer review automation should look like🧵👇
ALT First page of the ICML 2026 spotlight paper "Stop Automating Peer Review Without Rigorous Evaluation" by Joachim Baumann, Jiaxin Pei, Sanmi Koyejo, and Dirk Hovy (Stanford University and Bocconi University). The abstract argues that today's AI systems should not be used to produce paper reviews, grounded in two empirical findings: a "hivemind effect" where AI reviewers show excessive agreement and reduce perspective diversity, and "paper laundering," where prompting an LLM to rewrite a paper trivially increases AI reviewer scores through stylistic changes rather than scientific improvements. The paper calls for a science of peer review automation rather than wholesale deployment of general-purpose LLMs.
Tyler Menzer retweeted
When I talk about cheap travel, people really don't understand the middle class just didn't travel back in the day (or maybe their "travel" would be to a town or parkland or highway rest area two hours down the road by car)
I usually use airfares but let's use TRAIN fare in 1950
Tyler Menzer retweeted
My view on the California wealth tax proposal is probably more nuanced than some of my followers may want it to be.
Prop 40, as designed, has both strong pros and strong cons.
My main complaint is that the scientific debate around it has been polluted by political prejudices.
🧵
Absolutely thrilled for Axiom to finally tell us which utility function is the correct specification!
Tyler Menzer retweeted
I have been critical of some of the estimates, but the tax will absolutely collect less than expected, simply because the law appears like it will undervalue many private firms
Under the formula in the law, I would guess anthropic would be valued at ~$0
xitter.cf/JeffLHoopes/status/211…
The academic revenue estimates are literally just the Forbes numbers * .9 * .05. The wealth is just what Forbes says it is.
I think the GAAP point one is important--we have thousands of papers about how a GAAP number can be manipulated: taxpolicynetwork.org/when-we…
Some of discussion around this news is weird. Meta and Zuckerburg are not really distinct units. Zuck has majority voting control!
This wasn't Meta doing a sober NPV of Zuck being in CA, it was Zuck saying he wanted to stay and pulling the money out of his company to do it
The 50% legal problems haircut here is even worse that I thought. We should not apply a corporate accounting standard ad hoc to revenue estimates of bills, there is no reason for it. Imagine if CBO just halved the estimates of any bill because there could be legal challenges
