[ExI] The AI Superforecasters Are Here

[Adrian Tymes]

On Fri, Jul 3, 2026 at 8:54 AM BillK via extropy-chat
<extropy-chat at lists.extropy.org> wrote:
> On Fri, 3 Jul 2026 at 11:49, John Clark via extropy-chat
> <extropy-chat at lists.extropy.org> wrote:
> > The AI Superforecasters Are Here
> >
> > The following is a quote from the above:
> > "If the trendline does keep going, things start changing quickly. Finance gets transformed first, as human stock analysts go the way of horse-drawn carriages and kerosene lamps. The opportunity for smart humans to consistently make money on prediction markets likewise dries up - instead, bots duel other bots for the privilege of collecting money from dumb sports fans. [...] I asked the AI superforecasters the probability of a US-China treaty to slow down AI, enforced by cryptographic verification of data center activity. FutureSearch said 1%; Preseen, 2.2%."
> > _______________________________________________
>
>
> The article predicts that within a year, these specialised forecaster
> AIs will far surpass human forecasters. And for most cases, an
> off-the-shelf chatbot will be better at forecasting than most humans.
> This is partly because many human forecasters "talk their book". i.e.,
> they are trying to sell you something, which biases their forecasting.
>
> I liked the remark that "Every day, I see smart, tech-savvy people on
> Twitter voice opinions which a moment’s consultation with an AI  would
> reveal to be definitely false and stupid. We all have geniuses in our
> pocket willing to advise us on everything."
> Definitely worth remembering!

I wish...

On several topics, I have found AI consultation to be helpful.  On
finance, though, every time I've tried AI forecasting vs. what the
espert financiers were doing over the past few years, the AI lost.
(Good thing I only put it in test mode and didn't put actual money
behind it.)

I would not be too surprised if there are AIs that can and do reliably
make money on the financial markets today.  Just, apparently not any
AI that I have access to.

If I could gain access to such an AI that could predict to high enough
precision, even over relatively short timescales, I know where I'd
likely employ it.