Ed Zitron claims to have exact monthly cash spend (net of discounts and credits) by Anthropic on AWS. No idea the veracity. Since cash spend, it wont directly map to revs, but wth. Means Anthro is a 3.8% customer and driving 310bps in Y/Y growth in Q3 25 from $1.2B in cash spend. There are reasons for skepticism. Semianalysis showed that GCP was providing significantly more MWs to Anthropic than AWS up until this quarter, making the 2024 nums almost entirely reliant on training.

Two very interesting articles filled with unusually specific details about Israeli spying activities on Hezbollah printing in separate papers the same day. The why here as interesting as the what.

The FT in particular goes into greater details sourced from Israeli officials

Barclay's:

"the street is modeling $167B in cumulative AI capex, which is enough to support over 12,000 ChatGPTs.

We think one of the big players may blink and cut back the capex plans, but not likely until we get well into 2025 or beyond" "Based on these estimates, Google is assuming around 180T AI text queries (both input and output) and 15T AI image queries. This is a staggering figure, as there are around 11T web search queries per year right now worldwide. Stated differently, Google’s AI capex assumes a market that is 15x-20x larger than the web search market by 2026"

Reading the Nadella/Scott interviews, and the transcript from JPM of Alysa Taylor who heads Commercial Cloud GTM, you get insights on 3 key topics:

1) frontier models vs models-as-a-service
2) confidence in demand relative to capex spend
3) MSFT's attempts at differentiation Asked about vertical integration in AI, Nadella says "I'm more of the believer in the horizontal specialization".

More importantly, he goes on: "So I think any enterprise application, really what they're most excited about is models-as-a-service".

Something that was noticeable, on each of MSFT peer review slides for Apple, Amazon and Google, they highlighted progress on proprietary chips



And then the Azure and Windows slides both have develop custom silicon chips as long term drivers. We knew that this was the case, but just interesting to see them highlight other co's successes and set goals for it.

the TL,DR on the MSFT Board deck

Snowflake? Point player.

FTC request for comments on hyperscalers interesting. Can tell a lot by each participants responses.

ORCL/NET: whining about egress fees
GOOG: whining about MSFT/ORCL leveraging on prem biz to win cloud
MSFT: "very competitive market, even IBM"
AWS: history and tutorial on cloud this is arguably more descriptive from AWS then anything they've said to investors about AI

New collection of Messi vs Ronaldo charts. It’s pretty amazing how statistically similar they are. Ronaldo slightly more goals, but because of penalties. Messi more run of play goals. Messi far more assists, mostly after Ronaldo worried about injuries and stopped moving as much. Great chart contextualizing Messi and Ronaldo’s goal scoring

Look forward to someone doing the math but re AMZN wireless, are they gonna subsidize phones? Looking at Prime subs spend vs Non-prime, churn, and offering $0-10 lines, phone subsidy dramatically impacts net economics. If no subsidy, effective cost to consumers is much higher. Other calcs

- likely at risk to churn Prime HHs have lower spend than avg Prime HHs

- need to model ratio of at risk Prime HHs retained vs not at risk choosing wireless/reducing NPV

- as acq tool for non Prime HHs, almost certainly have lower spend/LTVs than existing Prime HHs

After a convo with smart friend, I wanted to test hypothesis that shift from Cable Nets to DTC meant more capital employed, smaller profit pool, and lower returns.

Very rough cut, but I took the media biz (Cable Nets/Studios, no Parks or O&Os except NBCU).

Results seem clear. To start, I disaggregated 2016 media co's into Cable Nets and Studio to the extent possible. What you can see is that the Cable Nets earned extraordinary returns on tangible assets. These were amazing biz. Studios ex DIS were not nearly as good. On the right are the consolidated.

If you're a business that makes money from monetizing the content creation of business executives or entertainers, charging them $8 a month is objectively one of the worst strategic decisions in the world.

If you get (got!) 2 US mDAUs, or 5 global ones, it pays for itself.

https://twitter.com/Jason/status/1649422446339006479

I have yelled this into the wind for years, Twitter does not have a monetization problem. They monetize at a high level all things considered given their total lack of direct response. Instead they have a user issue. They are trying to "solve" if that's the word, the wrong issue.

Was curious to hear Gary Gorton's thoughts on SVB, and he does not disappoint. Argues transactional deposits should be insured. But also argues Banks assets should only be known to Bank examiners because if depositors didn't know what SVB owned, the run wouldn't have occurred (!) Gorton also reiterates his view that financial crises are all about short term debt, and in this case the short term debt at root of the crisis was uninsured deposits. Says moral hazard is not a thing. And is very very worried about stablecoins.

there's a lot of political economy in this tweet

https://twitter.com/crampell/status/1640374330721796097

the government spent a bunch of money to prop up the economy and bridge households through a pandemic and... consumers don't remember getting the money or not losing jobs... but do remember the inflation that resulted. fascinating and has meaningful political ramifications IMO.

GS heavily caveat it and have all sorts of assumptions involved, but I will bet that generative AI comes nowhere close to raising baseline productivity by 150bps per year. The analysts must have gotten a homework assignment because three separate reports on AI hit tonight

"The largest U.S. banks didn’t submit a bid for Silicon Valley Bank over the weekend, largely because they were initially excluded from the sales process by the Federal Deposit Insurance Corp."

the first article to point blank say what has been rumored

semafor.com/article/03/14/… In other news, the NYT begins their weekend tick tock with an anecdote that has not appeared in any other story.

Deputy Treasury Secretary happened to take a meeting with Jamie Dimon.

Zuck with an open letter to every company in tech:

"Since we reduced our workforce last year, one surprising result is that many things have gone faster"

"Indirect costs compound and it’s easy to underestimate them"

"A leaner org will execute its highest priorities faster" Zuck "early analysis of performance data suggests engineers who joined in-person and transferred to remote or remained in-person performed better than people who joined remotely... engineers earlier in their career perform better when work in-person at least three days a week"

There is no need for a bailout or the Fed to “do anything” or any of the other things people are yelling for. All that needs to happen is the political will to allow one of the large banks to do what they would do as rational economic actors if they weren’t politically toxic. I do not understand the politics in this country well but since there will be no government funds at risk and no need for a backstop and shareholders are wiped out, I’m curious what the rationale for #OccupySandHillRd would be if JPM or one of the other money centers takes over.

struggling to see how this isn't a complete disaster if a real bank doesn't step in and assume the biz over the weekend. doesn't matter if uninsured are ultimately made whole, given the assets hard to see how it happens any time soon. curious for alt takes tho? i assume there will immediately be a market for the uninsured certificates, but let's say those trade at .50-60 (just making it up, but seems reasonable if recovery .85-100 but over years) runways just got cut in half.

“A new survey of NFL players presents an unusually revealing look inside how franchises worth billions of dollars are still rankled by problems—and found the Washington Commanders had the worst workplace conditions” Life imitates Major League:

“complaints included small hot and cold tubs for recovery, cramped locker room, lack of warm water and poor drainage in showers

Commanders are 1 of 7 teams that don’t offer first-class seats. also make segment of their players bunk with roommates”

Investors expecting management to get capital allocation exactly right is like LPs expecting investors to outperform all the time. Instead, investors should want management to have a capital allocation framework, just like LPs should want investors to have an investment process. In speaking of capital allocation, I’m assuming that the we get the platitude from every management team out of the way that “we first seek to deploy capital against internal investments”. It’s the question of what is done with the residual that is of great interest.

Dating myself, but its Bidz. com and its not even close. A no reserve auction site with mostly counterfeit jewelry and high end watches and the supposed main supplier was a convicted jewel fence. Put a 50 ct canary diamond up for bid. Almost certainly was a money laundering op.

https://twitter.com/CapitalCap1tal/status/1628934815272644609

Apparently there were tens of people willing to bid half a million dollars site unseen for a total unicorn of a stone