In partnership with
Welcome back and hello from one of my favorite places in the world — Las Vegas. I’m here to moderate two fireside chats at the Money20/20 fintech conference. My interviews from the stage with David Marcus, CEO of Lightspark, and Tony Huang, CEO of Possible, will drop next week.
This edition unpacks the growing trend of workers and companies selling their everyday work data to train AI models.
Plus, the latest inflation figures, CoreWeave stock’s very bad week, and big bank earnings tell us all the same thing: there are no signs of costs coming down.
And something new: A Photo of the Week.
Forwarded this email? Sign up here and support independent journalism. Send me feedback by hitting reply to this email.
Upgrade your free subscription to unlock the full Macro Talk experience.
The Rising Tide:
Feeding The Machines
Every week, The Rising Tide explores a trend that will shift the economy and the future of work. Today’s tide has a lot to do with a 1999 movie from Warner Bros.
What if I just used stills from “The Matrix” for every AI story I write? Maybe I should have saved this one for a future piece on agentic AI. (Photo: Ronald Siemoneit/Sygma/Sygma via Getty Images)
Would you upload your expertise to train AI models? For a couple hundred dollars an hour, more and more people are doing just that. But even if you prefer not to, your employer may already be doing it without your knowledge.
This week from Bloomberg:
As AI models improve, recruiters are searching for software engineers (pay: up to $120 per hour), condensed matter experts ($170 per hour) and civil rights attorneys ($400 per hour), all to “help train next-generation AI systems.”
Known in the industry as “experts,” these workers share the data economy with lower-paid generalists performing less specialized tasks.
The threat to workers materializes in two waves:
This becomes a new form of outsourcing, Melbourne Law School research fellow Fan Yang tells Bloomberg.
And the work of a few people could in theory replace or significantly devalue a whole field, says MIT professor Danielle Li.
The opportunity for corporates:
New revenue. Companies can package their specific set of knowledge work from their employees for wholesale.
“In such a future, almost everyone working with a computer could be considered a data worker, whether they like it or not,” Meaker writes in her piece.
But that future is already here — particularly for struggling startups.
SimpleClosure, for example, is a company that now helps founders sell their code and work data to foundation AI labs when their startups fail, MIT Tech Review reports.
Its CEO Dori Yona tells the outlet what was once “somewhat taboo” has become increasingly common — with nearly half of its deals including code or data sales.
What kinda data are we talking about?
Anonymized Slack DMs, emails with customers, Google Docs, PowerPoints, service tickets, past legal advice — you name it.
Micro1, an AI data startup valued at $4 billion, launched a program this year that pays midsize companies $100,000 to $2 million for access to anonymized data, per Business Insider. That access is valuable because AI models can train on longer time horizon data and get a fuller picture of how work gets done.
Business data has become the new oil. …
[A]ny company — dead or alive, big or small — can sell its data, as long as it has enough.
However, there are plenty of risks that executives and boards should consider, writes Daniel Goldberg, partner at Frankfurt Kurnit:
Legal: “A company that holds data does not necessarily have the right to sell it,” and “must comply with the privacy statements in effect when the data was collected.”
Goodwill: “Selling company data can damage trust.”
Competition: The quick cash grab today could lead to an AI company building a comparable product tomorrow.
What workers should think through
Your limited protections: “A national comprehensive data protection law does not exist, and most experts say such a law is not realistically on the horizon,” Helena Oroz, former senior legal editor at Brightmine, has noted.
Your value: Maybe there’s a figure you have in mind that would make the trade-off in training AI worth it.
Your desire to press management: Asking questions about what’s being done with your data is perfectly reasonable. How much you want to stick out is also part of the calculus.
A giant vacuum cleaner of the mind
For this elder millennial, I have the perfect dystopian analogy for this demand of our human output: Neo in “The Matrix” waking up in his pod.
I’ll never forget the first time I saw those tubes dislodge from his body. Are we not feeding the machines, in a way, at this point?
You can call me dramatic. But spending on data for AI could reach $6.5 billion next year — up from less than $1 billion last year — and most workers have no idea if their digitized labor will be part of that.
Tell me that’s not “The Matrix.” What do you think?
More Rising Tides:
A message from our sponsor:
Your AI prompts are leaving out 80% of what you're thinking.
When you type a prompt, you summarize. When you speak one, you explain. Wispr Flow captures your full reasoning (constraints, edge cases, examples, tone) and turns it into clean, structured text you paste into ChatGPT, Claude, or any AI tool. The difference shows up immediately. More context in, fewer follow-ups out.
Used by teams at Microsoft, Amazon, and Notion. Try Wispr Flow free on Mac, Windows, iPhone, and Android.
Decoding:
CPI and PPI, CoreWeave, Bank Earnings
There were three big data stories this week, so I’m going to hit all of them briefly because they’re all sending up the same flare.
No. 1: Energy costs rage
A 9.1% increase in gas prices made up more than half of the spike in the consumer price index (CPI) in September, Reuters reports.
Businesses, on the other hand, paid 20.3% more for gasoline, which accounted for about half of the jump in what producers charged for finished goods, per Barron’s.
Translation: Inflation is getting worse as energy prices remain high, TradeStation global head of market strategy David Russell writes.
No. 2: CoreWeave’s stock and the AI trade
CoreWeave’s 25% drop in share price this week underscores how sensitive the AI trade is to rising bond yields (what people want in exchange for holding certain debt).
If you’re unfamiliar, the company is a so-called neocloud — or an AI-first cloud computing provider.
And it’s “considered one of the riskiest stocks in the AI trade because of its heavy reliance on high-yield debt to finance its data centers,” Bloomberg notes.
No. 3: Big Banks kick off earnings season
It’s honestly always earnings season, but officially, banks kick off the cycle.
Results out this week from the country’s biggest names — JPMorgan, Wells Fargo, Citigroup, Goldman Sachs, Morgan Stanley and Bank of America — carried one central theme: warnings about higher costs for all stakeholders in the economy.
Photo Of The Week:
Economic Protests In Europe —
“Like A War Zone”
PARIS, FRANCE - OCTOBER 17: A car is set ablaze during demonstrations on October 17, 2026 in Paris, France. (Photo: Carl Court/Getty Images)
Peaceful protests over inequality and the high cost of living have spread across Europe and turned violent in some cities.
In France — where the first protests began in September with high school students demonstrating against public education conditions — more than half a million to some 800,000 people demonstrated on Saturday in marches. But masked rioters also looted businesses and set fire to buildings and cars in major cities like Paris.
"It's like a war zone here!" Dutch entrepreneur Tije De Boer told Reuters. He was visiting the French capital with his children and had to evacuate his hotel at Place de la République.
Belgium, Portugal, Italy, Spain and Albania have also seen protests arise and in some places, grow angry.
In the U.S., as consumer sentiment continues to worsen, the economy and the cost of living have become the top issues in the midterm elections.
Macro Talk will have more midterm election coverage in coming weeks. Send me suggestions and feedback by hitting reply.
Thanks for being a paid supporter!
A reminder of some upcoming programming and events from Macro Talk:
I’m hosting a dinner for CFOs in New York City on Oct. 28 and Dallas on Nov. 4. Shoot me an email if you know someone who would like to attend.
I’ll be at Money20/20 interviewing the CEO of Possible Finance Tony Huang and Lightspark CEO David Marcus.
In November, I’ll be headed back to Lisbon for Web Summit, and sitting down for two talks — Upwork CEO and president Hayden Brown, and a panel with OneStream CEO Tom Shea and Luminance CEO Eleanor Lightbody.
Forwarded this email? Subscribe your C-suite and board to Macro Talk here.
