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

Gary Marcus Marcus on AI Published

An open letter to Fei-Fei Li concerning California’s proposed AI regulation, SB-1047

We recorded 6 claims from this post. 4 were checked against evidence. 2 can’t be checked. 2 more are held. Read the post at its source.

Claims from this post (6)

In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.

  1. Assessment: SupportedSaid

    “Even then developers are required only to implement "reasonable safeguards" against these severe risks--not to fully mitigate them.”

    What we tested: SB-1047 requires developers only to implement reasonable safeguards against critical harms, not to fully mitigate them.

    What happened: The July 3 analysis describes reasonable safeguards and a prohibition on release where unreasonable critical-harm risk remains. Together these support risk reduction without a zero-risk guarantee. (Foley & Lardner LLP, )

  2. Assessment: Can’t be checkedSaid

    “Those seem reasonable to me, and I don't understand what would justify an exemption there.”

    What we tested: Marcus finds it reasonable for SB-1047 to focus on specific critical harms, and he does not understand what would justify an exemption from those provisions.

  3. Assessment: SupportedSaid

    “And much of what would be required is already something companies committed to voluntarily, in discussions at the White House and in Seoul.”

    What we tested: Much of what SB-1047 would require is something AI companies already voluntarily committed to at the White House and in Seoul.

    What happened: The announcement records pre-release testing, cybersecurity safeguards and sharing risk information. These overlap with central safety functions identified in the bill analysis. (The White House, announcing participating companies' commitments, )

  4. Assessment: SupportedSaid

    “But you don't make clear that much of the bill's requirements are limited to models with training runs of $100 million+. Companies that can afford that, presumably valued in the billions, are not exactly "little-tech."”

    What we tested: Much of SB-1047's requirements are limited to models with training runs of $100 million or more, and companies that can afford such runs, presumably valued in the billions, are not exactly little tech.

    What happened: The article expressly discusses the July 3 amendment and confirms that the $100 million training-cost threshold remained central. Its current page also contains a later veto notice, which is not used here. (Foley & Lardner LLP, )

  5. Assessment: Can’t be checkedSaid

    “With no other serious proposal on offer, I personally favor SB-1047, though I would welcome discussion of alternatives.”

    What we tested: With no other serious proposal on offer, Marcus personally favours SB-1047, though he would welcome discussion of alternatives.

  6. Assessment: SupportedSaid

    “Lastly, asking for standards is not unique to AI; it's common across many industries to ask that companies evaluate the safety of their products according to set standards: just look at the pharmaceutical industry, or aviation, automobiles, etc.”

    What we tested: Asking for standards is not unique to AI: in many industries, such as pharmaceuticals, aviation and automobiles, it is common to ask companies to evaluate the safety of their products against set standards.

    What happened: FDA describes manufacturers testing drugs and submitting evidence against a defined safety-and-effectiveness requirement. The event date denotes this standing regulatory practice at the post date, not the document's publication. (U.S. Food and Drug Administration, )

Held from this post (2)

These records did not pass every check, so nothing about them is published. Held is not a finding.

  • The two reviews reached different assessments. Our reading, not assessed: SB-1047 does not require AI developers to predict every possible use of their models.
  • The criteria are ambiguous. Our reading, not assessed: As Marcus understands the latest version of SB-1047, its kill switch requirement does not apply to open-source models once they are out of the original developer's control.