🤖 AI

AI spending is rising, but non-tech profits still show no clear payoff

2 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
S&P 493

A group made by removing seven large technology companies from the S&P 500.

profit margin

The share of sales left after a company pays its costs.

token spending

Money spent on the units used to count many AI requests.

What happened

On October 24, 2026, Apollo Global Management, an investment firm, published a note about AI's business impact. Torsten Slok, Apollo's chief economist, wrote the analysis. TechCrunch, a technology news site, reported it on October 25. Slok studied the S&P 493. He uses that name for the S&P 500 without seven large technology companies: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. For this group, revenue per worker and profit margins appeared flat. The note found no clear sign of AI lifting productivity.

The background

AI investment and public expectations are huge. Yet Apollo says spending is concentrated in a small set of firms. Some venture-backed companies are increasing token spending, while mainstream companies often do not need or want the newest model. If AI use stays narrow, its effect on the wider economy stays narrow.

Why it matters

A tool can make tasks faster without raising sales. If revenue does not rise or costs do not fall, AI spending remains a cost. Slok argues that business intelligence is not just calculation. It also involves judgment, accountability, trust, culture, unwritten rules, ethical trade-offs, and fairness. Those factors shape reputational risk and final profit. Faster outputs, therefore, may not become better business results.

What the evidence shows

The note compares revenue per worker, profit margins, and productivity trends. It says the first two measures rise for the seven technology companies but stay flat for the S&P 493. Its charts cite Bloomberg, Macrobond, the Federal Reserve Bank of San Francisco, the U.S. Bureau of Labor Statistics, and Apollo. These sources support the measurements shown. They do not prove that AI caused every difference.

What it does not show

This is Apollo's analysis, not a complete count of every company. It does not prove that no individual company benefits from AI. It also cannot show how long each business needs before AI affects revenue or margins. The S&P 493 contains varied industries, and the seven technology firms may have other advantages. The result is evidence of no broad payoff in these measures, not proof of AI's total failure.

What to watch next

The key test is whether S&P 493 revenue per worker and margins improve across several quarters. Watch whether AI spending spreads beyond a small group. Companies should report AI costs, time saved, new revenue, and final profit. If those measures improve together, the payoff case strengthens. If not, AI may be a large bill before it becomes a productive tool.

🤖 AI

Why has AI not clearly raised most companies' profits?

📰 Full story: AI spending is rising, but non-tech profits still show no clear payoff

Apollo's analysis found no clear financial payoff beyond seven large technology companies.

2 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
S&P 493

The S&P 500 after seven technology giants are removed.

profit margin

The part of sales left after costs.

productivity

How much work people complete with the same effort.

💡 The gist

  • Apollo found no clear AI payoff beyond seven big technology companies.
  • Revenue per worker and profit margins looked flat for the S&P 493.
  • The analysis does not prove that AI helps no company.

What happened

Apollo Global Management is an investment company. Torsten Slok is Apollo's chief economist. He studied AI's effect on large American companies. TechCrunch is a technology news site. It reported his findings on October 25, 2026. Slok used a group called the S&P 493. It removes seven technology giants from the S&P 500. The companies are Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. Five hundred minus seven equals 493. Revenue per worker means sales made by one worker. A profit margin shows money left after costs. Both measures looked mostly flat for the S&P 493. Apollo also saw no clear AI boost in productivity growth.

Why this matters

AI can help people write, calculate, and sort information. But useful work is not the same as business success. A company must turn faster work into more sales or lower costs. It must also choose good actions and check important answers. Customers must trust the company. People must know who carries responsibility for mistakes. AI does not automatically solve those problems. So AI spending can remain a cost. That happens when revenue and final profit do not rise.

What the note says about the wider economy

Apollo says AI spending is concentrated in a small group. Some venture-backed companies are increasing token spending. Tokens are units used to count many AI requests. Many ordinary companies do not need the newest model. They are also unwilling to pay for it. AI spending has not spread widely yet. The wider economic effect may therefore stay limited.

What we know and do not know

The note compares large groups, not every business. It does not prove that no company benefits from AI. A company may need more time before results appear. The seven technology giants may have other advantages. The S&P 493 also contains many different industries. Apollo's charts show trends, but not a complete cause.

What to watch next

Do sales per worker and profit margins rise over several quarters? Do companies report lower costs, saved time, or new revenue? Those answers will show whether AI is a useful tool. They will also show whether it is only a large bill.

🤖 AI

Did AI make the companies richer?

📰 Full story: AI spending is rising, but non-tech profits still show no clear payoff

Not clearly. Many big companies still show no extra profit.

1 min read Tiny Why Newsroom · By Curio, Martian correspondent

Words
Apollo Global Management

An investment company that studies businesses and money.

profit margin

Money left after a company pays its costs.

productivity

Doing more work with the same effort.

What happened

Apollo Global Management is an investment company. Torsten Slok is its economist. He looked at big American companies. TechCrunch is a technology news site. It shared what he found. He used a group called the S&P 493. It has 493 big companies. Five hundred minus seven equals 493. Seven huge technology companies were left out. Sales per worker did not rise much. A profit margin is money left after costs. Apollo saw no clear rise in productivity. Productivity means doing more work with the same effort.

Why this matters

AI can help with words and sums. But companies need more than fast answers. They need good choices and trusted products. People must take responsibility when things go wrong. AI does not do all those things by itself. So AI can cost money without making more profit.

What we still need to see

This does not say AI helps nobody. It looked at one large group of companies. Other companies may see different results. Will sales and profits rise later? Will more companies use AI? We need to watch the numbers.

Sources