Microsoft’s big stick era
And what is Sam Altman talking about?
Welcome to Cautious Optimism, a newsletter on tech, business, and power.
Happy Thursday, friends. This morning we’re all watching the Starship explosion that happened late last night. The cause is not yet nailed down, and no one got hurt — but the anomaly during prep for a reported static fire test is a setback for SpaceX. Which means that the future date at which I will be able to afford to go to space just got pushed back.
Yesterday the Fed did not cut rates (as expected), leading to mean tweets (as forecasted). The Bank of England ran the same playbook. Stocks were down in Asia, and down in Europe. American public markets are closed today to respect the Juneteenth holiday. (A good Juneteenth explainer can be found here.) With all that in hand, to work! — Alex
📈 Trending Up: Non-equity venture capital … sports profits … free trade polling … crypto hacks as war tools … media dealmaking? …
📉 Trending Down: EU deficits … de-escalation between Iran and Israel … Samsung’s chip prowess … the 737 Max …
📝Paper of the Day: This MIT paper compares brain activity differences between people who use AI tools to help with cognitive work, and those who do not. It also ran tests on folks who initially used AI in its tests and no tools later, and those who initially used no tools and later were given access to AI tools. The results? “The use of LLM had a measurable impact on participants, and while the benefits were initially apparent, as we demonstrated over the course of 4 months, the LLM group's participants performed worse than their counterparts in the [no AI tools] group at all levels: neural, linguistic, scoring.”
Microsoft’s big stick era
After cutting around 6,000 jobs — around 3% of its total staffing — back in May, Microsoft is reportedly prepping another RIF. This time, Bloomberg writes, the “thousands” of roles that will get deleted will center around sales staff.
We’re now accustomed to the largest, and wealthiest tech shops regularly curtailing their human capital. (Here’s Google in June, Amazon in June, Meta in May.) But Microsoft’s continued reductions are starting to feel a bit more than cosmetic.
The company is certainly in rude health. Total revenue growth in its most recent quarter came in at 13%, accompanied by even larger gains in operating income (+16%), and net income (+19%). Hell, ‘Azure and other cloud services’ grew 33% in the period.
Redmond is so profitable that its stated $9.7 billion worth of shareholder return in the third quarter of its fiscal 2025 (calendar Q1 2025), the company spent over $107 million per day on its shareholders. (Of which, it’s worth pointing out, is all of us who own an index fund.)
Microsoft is secure enough, then, that it can trim costs while growing, scaling its incomes, and dropping gold on its shareholders. The company’s oats are sufficiently strong that it’s also content to play a little hardball with OpenAI. The FT reports that Microsoft is prepped to walk away from talks with the AI giant concerning its corporate makeover. That would freeze the relationship in its current form, annoy lots of OpenAI backers, and set the stage for a divorce in 2030, when the two companies’ agreement runs its course.
I doubt that that will happen, but Microsoft appears unconcerned if things do go south with is AI partner. You know what they say, Never go in against a Sicilian Satya when death cloud profits are on the line.
AI job destruction
The Andy Jassy AI memo made it clear that some large companies expect their overall staffing needs to decline over time thanks to new software tooling. It’s not a hard point to understand:
AI is/will make workers more productive;
Therefore, companies will need fewer human in the future to accomplish required work over time.
AI tools will also generate work, but net-net Jassy expects Amazon to do more in the future with fewer humans on salary. While that memo was circulating, my co-host on TWiST Jason Calacanis got to ask Jassy a few questions. Here’s what he learned, pulling from yesterday’s show (condensed slightly, emphasis added):
My first [question was about AI-led] job destruction […] You’ve seen the Figure robot sorting things, you have tons of robots in your factory, self-driving, you're doing Zoox. It seems to me that when a lot of those jobs go away, and obviously you're not hiring a lot of people, you have less people at Amazon than you did a couple years ago, so it's happening in white collar. And he was basically very upfront and candid that he did think there was going to be a lot less people working at companies and that it was something society would have to navigate. And that is the tension. […]
I think we have some responsibility. I don't buy into, like, technology is just good at technology and let the chips fall where they may. I think we actually should think about it. I know it's a bummer, but we should think that, hey, some jobs will be disrupted.
Now, is it our responsibility to create all the new jobs in the world? No. Of course not. But I think we should just be thinking it as human beings on planet Earth because there could be some disruptions.
I use Jason as something akin to a touchstone for internal technology industry sentiment. He’s in the group chats that I am not, and spends time talking to other investors in rooms where I am very much not invited. Welcome to being part of the press.
His expression of concern is really notable, then. You might have expected the capitalists — my team, mind — to be a bit more callous. Not so.
One more nibble: Jason went on to argue that one current employment safety net, gig work for Uber and similar, could also get automated away. Sidewalk robots, self-driving cars, and drones could replace some casual human jobs. Put another way, AI is automating the union factory job, the gig worker’s next order, and white collar work that college-age kids dream of. That’s broad labor pressure.
What is Sam Altman talking about?
There’s a clip from the new OpenAI podcast — of course — concerning xAI’s Elon Musk and his time in government. The Sam Altman quote that has people talking goes as follows:
I had said, I think also externally, but at least internally after the election that I didn't think Elon was going to abuse his power in the government to unfairly compete. And I regret to say I was wrong about that.
I mean, I don't like being wrong in general, but mostly I just think it's really unfortunate for the country that he would do these things, and I didn't think I genuinely didn't think he was going to. I'm grateful that the administration has really done the right thing and stuck up to that kind of behavior.
This is not a little intra-technology spat. Making the claim that a rival used government power to try and bias national policy against OpenAI is a big deal.
The context here is actually included in the longform podcast, but since there are more podcasts out than people have time to listen to — excepting TWiST, which I trust you tune into three times per week — I can supply the asked question:
I read a behind the scenes story about the development of Project Stargate and the international partnerships, particularly the UAE, and that Elon Musk had tried to derail that. And what have you seen? What have you heard? What's the take on that?
Let me explain. POTUS went to the Middle East back in May. During the trip the UAE announced big datacenter plans. The first phase of the AI-focused compute nexus will consume one gigawatt of energy, slated to eventually scale to five gigawatts of power consumption.
For reference, the big French datacenter that the UAE is also involved with is slated to cap out at around the one gigawatt mark.
OpenAI’s involvement in the project — part of its multinational Stargate compute push — was a big deal, capping off work behind the scenes to get multiple governments on the same page. The United States had to work through concerns that partner in the project, the UAE-headquartered G42 was too friendly with China. The UAE eventually earned access to purchase gobs of Nvidia chips, and the influence race between China and the United States when it comes to Middle Eastern political and economic leanings tipped a bit more in America’s favor.
But in the background of the various deals, investment promises, and general backslapping was a sticking point: Musk wanted xAI to take part in the overall agreement. The Journal reported the following:
On a call with officials at G42, an AI firm controlled by the brother of the United Arab Emirates’ president, Musk had a warning for those assembled: Their plan had no chance of President Trump signing off on it unless his company xAI was included in the deal, according to some of the people.
Musk had learned just before Trump’s mid-May tour of three Gulf countries that OpenAI Chief Executive Sam Altman was going to be on the trip and that a deal in the U.A.E. was in the works, and grew angry about it, according to White House officials. He then said he would also join the trip, and appeared alongside the president in Saudi Arabia.
Altman’s reference to the second Trump administration not knuckling under seems to allege to the fact that despite alleged pressure from Musk, the deal went ahead sans xAI. Now, in a house podcast, Altman all but confirmed the reporting.
All is fair in love and war, but I think in business we can expect a bit better behavior than using government influence to try and bully investment decisions.
Cloudy with a chance of rate cuts
The Fed expects two rate cuts this year, good news for folks hoping that monetary policy will become more accommodative. Lower rates are good for higher-risk assets, and generally correlate to higher prices for technology shares.
Past the bluster, what does the Fed see in the US economy today?
Reasonably solid economic growth (“Private domestic final purchases, or PDFP as we call them —which excludes net exports, inventory investment, and government spending—grew at a solid 2.5 percent rate.”)
Slower future GDP expansion (“In our Summary of Economic Projections, the median participant projects GDP to rise 1.4 percent this year and 1.6 percent next year, somewhat slower than projected in March.”)
A balanced labor market (“Payroll job gains averaged 135 thousand per month over the past three months. The unemployment rate, at 4.2 percent, remains
low and has stayed in a narrow range for the past year. Wage growth has continued to moderate while still outpacing inflation.”)
Rising inflationary pressures (“Near-term measures of inflation expectations have moved up over recent months, as reflected in both market- and survey-based measures. Respondents to surveys of consumers, businesses, and professional forecasters point to tariffs as the driving factor.”)
That last point matters. During Q&A (transcript source), Fed Chair JPow further explained what he’s seeing regarding prices:
We’ve had goods inflation just moving up a bit. And, of course, we expect – as you point out, we do expect to see more of that over the course of the summer. It takes some time for tariffs to work their way through the chain of distribution to the end consumer. A good example of that would be goods being sold at retailers today may have been imported several months ago, before tariffs were imposed. So we’re beginning to see some effects. And we do expect to see more of them over coming months. We do also see price increases in some of the relevant categories like personal computers and audiovisual equipment and things like that that are attributable to tariff increases.
The Fed is therefore in a pickle. If it cuts rates to bolster slowing expected GDP growth just as inflation from tariffs starts to tick up, we could wind up in an environment with far greater price increases than accelerated economic expansion. Worse, if trade tensions and a hot geopolitical landscape hamper consumer and business spending as inflation rises, we could start to hear the unspoken word stagflation start to rejoin our daily conversation.
So the Fed is staying pat for now, watching the numbers, and waiting for the right moment to cut the price of money. POTUS is peeved, but as the chief requirement for central bank leadership appears to be patience, we shouldn’t be shocked.
For whatever it’s worth, if I was sitting in the Fed throne I would also wait for more tariff-related price changes to filter through the market before cutting rates. But I’m also too nitty at poker, so make of that what you will.
Tomorrow, provided the news cycle is kind, we’ll take a look at staffing levels at major cloud companies in the last few years to get a better feel for the aggregate scale of layoffs that we’ve seen in 2025, and run through a list of the startups I’m adding to the TWiST500. There’s some incredibly cool companies being built today. Onward!