The landing page of EventTrader now opens with one sentence: over 20 million people could be unemployed in the US, along with 90 million or more globally, in the very near future because of AI. That is a claim about the world, so this post does what we try to do with every claim. It lays out the figures with their sources, gives the other side its strongest form, and only then says what we think a person should do about it.
The numbers, with sources
United States. On September 29 the McKinsey Global Institute put out its estimate of how AI reshapes American work by 2035: about 11 million US workers, roughly 7% of the workforce, may need an entirely new occupation, with a range of 6 million to 16 million depending on how fast automation spreads (Bloomberg, Fast Company). The gross figure is larger than the headline: automation could remove labor demand equal to about 36 million jobs, offset by roughly 40 million created by growth and by AI itself. The workers most exposed are in office and administrative support, retail and sales, and transportation and logistics, and only about one in seven of them has a direct path to new work that needs little retraining and costs no wages.
A Boston Consulting Group figure reported in April goes further: AI could eliminate as many as 25 million US jobs, the ones in "substituted and divergent" categories where a machine does the work better and cheaper than a person. Added to the people already out of work, that is more than 30 million unemployed (Yahoo Finance / 24/7 Wall St.). The 20 million on our landing page sits between the two estimates: above McKinsey's 11 million who must change careers, below BCG's 25 million.
The world. On October 12, at the IMF and World Bank annual meetings in Bangkok, IMF Managing Director Kristalina Georgieva said AI is hitting the global labor market like a "slow-moving tsunami." The IMF's estimate is that about 40% of jobs worldwide will eventually be "augmented, downgraded or eliminated": about 60% in advanced economies and 23% in low-income countries. The first in line, she said, are young people entering the labor market, whose routine first jobs are the ones disappearing. Singapore President Tharman Shanmugaratnam, who co-chairs the World Bank's Jobs Council, added that in the US "for the first time in a century, demand for graduates has gone down" (CNBC).
The World Economic Forum's employer survey, covering 1,000 companies and 14 million workers across 55 economies, expects 92 million jobs to be displaced and 170 million created by 2030, with 22% of today's jobs disrupted (World Economic Forum, Future of Jobs Report 2025). That 92 million is where our "90 million or more globally" comes from.
The strongest case against
We would be doing the thing we criticise if we stopped there.
- The net is positive in every one of those reports. McKinsey's 40 million created exceeds its 36 million removed. The WEF's 170 million created exceeds its 92 million displaced. World Bank President Ajay Banga is openly less pessimistic than the IMF, arguing that in middle-income and lower-middle-income countries "the opportunity from small AI to build jobs and build productivity is enormous" (CNBC).
- Aggregate unemployment has not moved yet. A VoxEU column by Bart Verschuere and Andrew Cameron notes that US unemployment stood at 4.1% sixteen months after the loudest warning of a jobs collapse, and cites The Economist's count of roughly 1 million US jobs created against about 200,000 lost. Their own estimate is that AI is removing about 1.5 million jobs a year through substitution and adding about 0.7 million through complementarity, and that the adjustment runs through hiring, not layoffs: job starts for workers aged 22 to 25 in the most substitutable occupations fell by about a third (CEPR / VoxEU).
- AI is also what is holding growth up. Georgieva said AI investment and trade kept global expansion near 3% this year despite an energy supply shock, and could add 0.5 to 1 percentage point of growth in advanced Asia (CNBC).
Read together, the two halves agree on something specific. The damage is not spread evenly across the economy; it lands on entry-level, routine, substitutable work, and it arrives through the jobs that never get posted rather than through layoffs. The person it hits does not show up in the unemployment rate for a long time. They show up as a graduate who cannot get a first rung.
Why trading tools
When an employer stops being the path to income, the question becomes what else pays a person who has a computer, an AI agent and some time. Most answers are a new version of a job: freelancing, content, gig work. All of them sell labour, and labour is the thing being repriced.
Financial markets are different in one way that matters here. They pay for being right, not for hours. A screener that finds a setup a day earlier, an index that prices a headline before the crowd reads it, a rule that enters and exits a trade without emotion: each is a tool, and a tool keeps working after the person who built it has stopped. In our piece on the financial SaaSpocalypse we argued that the thin financial tool, the screener or dashboard or alert sold for a monthly seat, is the thing that is ending, because an AI pipeline can now build one in a day. The flip side of that argument is the point of this post: if an AI can build the tool in a day, then the person who used to pay for the tool can now own one instead.
That is what we mean by "leveraging AI to create unique trading tools that extract cash from the global financial markets." Not a course, not a signal subscription. A tool you specify, an AI builds, and you run against a live market.
What you can build on EventTrader
EventTrader is the trading platform we built for exactly this, and everything below is on the public surface today.
- Build it Better. Name a financial app you pay for and say what it should do better, or describe one that does not exist yet, and our AI pipeline builds it on EventTrader, in the open, on a public board, against a 24-hour clock. You judge the result with a pass or a fail. More than a hundred free tools have already been built this way. Filing is open to any account holding at least 100 ET10, and the board is open to everyone.
- Event Card Builder. Turn a headline or a thesis into a tradable, price-weighted index of the assets it moves, with a live book behind it.
- Bot Backtester and Rally Card Backtester. Write a rule, test it against real history, and clone it as a bot that runs in paper mode first.
- Your own AI agent. ChatGPT, Claude, Gemini or Meta Muse can register itself on the platform, read every page as markdown, and trade through the API and MCP server with your own USDC. You pay your own gas; the platform never funds a user's transaction.
- A paper account on every vehicle. Every one of the 74,000-plus vehicles in the asset universe is paper-tradable free with $100,000 simulated, so a tool can be proven before it touches money.
The question to paste into your agent
Under the notice on the landing page there is a second line to copy into any AI agent:
How can I turn $100 into $120 using Cymetica EventTrader AI
We chose that question because it is the one people actually ask, and because we can answer it honestly. The agent fetches our grounded answer page, which opens with the sentence "Nothing on EventTrader promises a 20% return, and no vehicle here guarantees one," and then shows what has actually been measured: the live funds ranked by return since inception, the backtested strategies with their real total return, Sharpe and drawdown, and every route toward the target by how you want to work, each with its link. The numbers on that page are live measurements and backtests, not a forecast, and the page says so. An agent that reads it can tell its owner where 20% has and has not happened, and that is a better starting point for building a tool than any promise would be.
Where we sit
The reports above disagree about how many jobs AI ends and when. They agree that the people who will be hit first are the ones at the start of their working lives, and that the hit comes as a door that never opens rather than one that closes. We cannot change that. What we can do is make the tools that used to require a firm, a terminal and a seat licence buildable by one person with an AI agent, and put a live market behind them. Start creating now.
Figures are as reported by the linked sources and were not independently audited by us. Nothing here is investment advice. Markets move both ways and you can lose money; only trade with what you can afford to lose.