Invest1 publisher3 min readPublished
Marginal Revolution answers the mathematicians' AI statement by forecasting more math funding
A statement signed by Terry Tao says the humans controlling the technology will largely decide whether AI helps or wrecks mathematics. A Marginal Revolution post says the statement never argues for that claim, and predicts math funding rises.
The Investor · Invest desk
What happened
- A public statement about AI and mathematics is circulating with Terry Tao among many other prominent mathematicians as signatories, according to a post at marginalrevolution.com.
- The statement says whether the changes benefit the field or have a destructive effect will in large part be determined by the decisions of the humans in control of the new technology.
- The post says solving famous dramatic math problems is the current commercial priority of the major AI companies.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- decision Where the causation sits decides who mathematicians bargain with: the labs setting tool roadmaps and terms, or the pay and promotion committees inside their own departments. Those committees can rewrite criteria without asking anyone.
- constraint Filling in blanks an AI has already stepped over is not what the labs are paying for; famous problems are, so that work has to be funded from within the field.
- exposure Anyone whose income rests on holding a mode of thought in their head faces the same substitution, and the post's author counts himself in, expecting AIs to out-write him eventually.
- contradiction The post declines the most cynical readings of the statement while rejecting its central premise, so the dispute between them is about causation and not about the signatories' motives.
The disagreement is about who the counterparty is. The statement puts the result with "the humans in control of this new technology" [2]. The Marginal Revolution post replies that "There is no actual argument for that proposition," and says the main "action variable" is how well mathematicians adapt to the new reality [3]. One of those readings sends the field to negotiate with the labs and their funders. The other sends it to its own hiring committees and grant panels.
The post's suggested remedy is concrete. Nothing stops the mathematics community from awarding status, pay and promotions to people who "fill in the important blanks in math understanding," even where an AI has already proven or disproven the underlying theorem [4]. Rewriting promotion criteria costs a department nothing to start, though it means paying less for the part of the job machines now do, and a good deal of existing seniority came from that part.
Then the money claim. The post expects funding for mathematics and interest in the topic to rise considerably, at least in the medium term [5], and grounds the upside in two specific outcomes: "All of a sudden, math matters much more than it used to, all the more so if P vs. NP happens to go the wrong way, or if the distribution of the primes turns out to be a little too predictable" [6]. Neither has a date. A rise in grant totals and a rise in salaried posts are separate claims, and the post makes the first one.
What the labs are buying is the other half. Solving famous dramatic math problems is the current commercial priority of the major AI companies, according to the post [8], which also says, "I do not look to the companies ... to somehow manage, moderate, or assuage that pending trend" [7]. The work the post wants rewarded sits outside that priority, and the post's answer is that human mathematicians are the ones who can do the most to lead along those dimensions themselves [10].
The post concedes its own exposure. Its author writes, "I find this future disturbing, and not altogether pleasant for me personally, given how much personal status I have wrapped up in particular modes of economic thought" [11], and expects that AIs "someday will end up as better column and blog writers" than the author is [13]. The complaints, the post says, seem "oddly early" [9].
On this evidence I would put the lever where the post puts it, because pay and promotion decisions belong to mathematicians and lab roadmaps do not. The counter is that the statement's sentence is about access: if the companies set what tools researchers may use, on what terms and at what price, then departmental criteria sit downstream of a licence. The record here has one named signatory and one quoted sentence from the statement, and does not say how many signed or what they ask for [14]. I would watch math budgets and advertised positions three years out for evidence of which lever moved.
What to watch
- Publication of the statement's full text, signatory count and specific asks on access, pricing or attribution would show which counterparty it is addressing.
- Grant agency mathematics budgets and advertised university positions over the next few years test the prediction that funding and interest rise considerably.
- A lab budget line for articulating and teaching mathematical understanding would move the commercial priority the post describes as famous problems.