Pundit ClaimsAI claims by Gary Marcus and Ed Zitron, with the evidence

“Any advantage in Generative AI (which still hasn’t remotely made a net profit, and is all still speculative) will be minimal, and short-lasting.”

When it Comes to AI Policy, Congress Shouldn’t Cut States off at the Knees, paragraph 9.

Our reading:

Generative AI has not remotely made a net profit and remains speculative.

Assessment: Unresolved

What happened

Was it true for the period described?

Outcome: Unresolved

We checked, and the evidence so far doesn’t settle it.

Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.

Evidence up to . Checked .

Why

Ars Technica reported on June 16, 2026 that OpenAI's net loss for calendar 2024 was just over $5 billion [e1]. Anthropic's October 26, 2026 filing reports a net loss of $8.3 billion for calendar 2024 [e2]. On August 1, 2022, The Register quoted Midjourney's founder saying the company was profitable, without specifying a net-income measure or accounting period [e3]. These company-level results and statements establish neither aggregate profitability nor its absence across generative AI through May 14, 2025, including cumulative results [e1] [e2] [e3].

Caveats

  • The evidence lacks comparable industry-wide net-income records and cumulative results through May 14, 2025.
  • The records do not isolate generative AI results within diversified companies or establish consistent shared-cost allocation.
  • Midjourney's profitability statement supplies no financial statements or defined accounting measure.
  • The claim leaves the accounting basis unspecified and gives no separate measurable test for 'speculative'.

Evidence (3 items)

  • Context only: Ars Technica,

    in 2024

    Why it matters: The 2024 result documents losses at one leading provider; the 2025 comparison is outside the claim's target period. This is reporting on audited statements, with Financial Times review identified, rather than an industry-wide account.

    Publisher
    Ars Technica
    Kind of source
    News report, audited
    Measures
    OpenAI, Reported annual net loss, USD billions, company reported, 2024-01-01 to 2024-12-31
    About events on
    Published
  • Context only: Anthropic, PBC / U.S. Securities and Exchange Commission,

    history of net losses

    Why it matters: The accompanying paragraph reports losses in each year since founding, including a $8.3 billion net loss for 2024. That historical company result is relevant context but cannot establish cumulative profitability for generative AI collectively.

    Publisher
    Anthropic, PBC / U.S. Securities and Exchange Commission
    Kind of source
    Regulatory filing
    Measures
    Anthropic, Annual net loss, USD billions, company reported, 2024-01-01 to 2024-12-31
    About events on
    Published
  • Context only: The Register,

    To be honest, we're already profitable, and we're fine.

    Why it matters: Midjourney's founder reported profitability before Marcus's statement, challenging a reading that no provider had made money. The statement does not identify net income, an accounting period or an accounting standard, and cannot establish industry-wide profitability.

    Publisher
    The Register
    Kind of source
    News report
    Measures
    Midjourney, Founder-described profitability; profit measure unspecified, company reported, ? to 2022-08-01
    About events on
    Published

Probably right?

One AI thinks this is probably right, and the other says it’s too early to tell. The evidence isn’t conclusive, so the verdict above stays unresolved. It counts only when you pick Speculative under “How sure?”.

  • Anthropic’s AI: OpenAI lost just over $5B in 2024 (reporting on audited statements) and Anthropic's SEC filing shows an $8.3B loss and losses every year. Midjourney's unaudited 2022 profit claim doesn't offset this.

How we judged it

Kind of claim
Observation
Deadline
None given. We don’t add one.

Where we had to interpret

  • Assess the parenthetical observation about profitability, not the surrounding forecast about national competitive advantages.
  • The ordinary reading is that generative AI as a commercial field had not achieved overall net profitability by May 14, 2025. This does not necessarily assert that every individual provider was unprofitable.
  • The words 'hasn’t ... made' encompass profitability achieved up to the statement, including cumulative results. The target period therefore ends on the post date and has no invented starting date; it is not restricted to calendar 2024 or the latest quarter.
  • Read 'remotely' as emphasis on the claimed absence of profitability, without inventing a numerical distance from break-even.
  • Read 'speculative' as describing the uncertain prospect of sustained commercial profits. It supplies no separate measurable threshold.
  • The speaker supplies no accounting basis, currency, industry boundary, or method for allocating shared costs. These cannot be supplied silently.

Right if: Comparable financial records or a transparent industry estimate establish that generative AI businesses collectively had not achieved net profitability by May 14, 2025, including cumulative results through that date. Losses at selected leading companies alone are insufficient.

Wrong if: Comparable financial records or a transparent industry estimate establish that generative AI businesses collectively had achieved net profitability by May 14, 2025. An individual profitable provider is relevant counterevidence but does not by itself establish aggregate profitability.

We wrote this test after the outcome was known.

What would settle it

Comparable financial records or a transparent industry-wide reconciliation of realized net income through May 14, 2025, including cumulative results and any earlier achievement of aggregate profitability. It must establish industry coverage, a justified accounting basis, shared-cost allocation and treatment of transactions between providers.

We found 18 claims in this post. See them all.

Full record

Source and dates

Marcus on AI, When it Comes to AI Policy, Congress Shouldn’t Cut States off at the Knees, paragraph 9.

Said
We read the source
live page, full text, public
Evidence up to
Checked

We read this wording on 2026-10-28. The post may have been edited after 2025-05-14.

Where it was said (1)

Said once in the posts we read.

  • : Marcus on AI, When it Comes to AI Policy, Congress Shouldn’t Cut States off at the Knees

    Any advantage in Generative AI (which still hasn’t remotely made a net profit, and is all still speculative) will be minimal, and short-lasting.

    Published
    2025-05-14
    We read it
    , live page
    Access
    full text, public

History

Versions

  • Current: Unresolved. Evidence up to 2026-10-28.

What happened to this record

  1. PublishedBy us

    Published after both AI reviews and the final check passed.

  2. Held backBy us

    A reviewer said our test wasn’t fair to what was actually said.