Will AI have a sudden trillion+ dollar impact by the end of 2023?
➕
Plus
86
Ṁ19k
resolved Jan 6
Resolved
NO

"Trillion+ dollar impact" is a proxy for "obviously life-changing for normal people". It need not count as human-level AI aka artificial general intelligence (AGI). Examples of things that would surely count:

  1. Virtual assistants that are better than well-paid humans

  2. Superhuman art; people tend to prefer to read / view / listen to AI-generated art

  3. AI generating wholly new science/tech

  4. Level 4 self-driving cars fairly ubiquitous

  5. Level 5 self-driving cars available in some places

  6. (A technological singularity is irrelevant for betting purposes)

In case it's hard to quantify the impact in dollars, the following is a necessary but not sufficient condition for this market resolving YES: I ask friends/family who pay zero attention to AI developments and they are extremely aware of the thing.

Get
Ṁ1,000
and
S3.00
Sort by:
predicted YES

Everyone agree on NO here? Closest was ChatGPT but OpenAI's valuation hasn't hit a trillion dollars (a necessary but not necessarily sufficient condition for this to resolve YES).

predicted NO

@dreev Agreed, we clearly aren't at the trillion+ impact yet.

Superhuman art; people tend to prefer to read / view / listen to AI-generated art

presumably this means 'people prefer AI art to the best existing human art'? Because people already prefer AI art to most casual human artists

@jacksonpolack I prefer the best existing AI art to most casual human art, but looking at my median result from Manifold Dream I prefer casual human art.

Eg, this is a lovely picture but doesn't look especially like a casual human artist to me.

casual human artist
predicted YES

To name another hypothetical, do people agree that if OpenAI were to hit a trillion dollar valuation, that that should count as a YES here?

predicted NO

I don't think that counts. Finance is time travel; the valuation of a company reflects its net present value extended indefinitely into the future. A company worth a trillion dollars today hasn't produced a trillion dollars, yet. It is expected to produce a trillion (time-discounted) dollars over its existence, but that's not the same as having produced a trillion dollars already.

That said, "a trillion dollar impact" has some wiggle room in it—you could say that a time-discounted EV of a trillion dollars is an "impact" of that magnitude. But: to the extent that the intent here is, as the description states, a proxy for "obviously life-changing for normal people", it's clear that an expectation of future impact (post-2023) shouldn't count. Given that that's how you've defined "impact" in the description, I think you can't resolve this on the expectation of future effects.

Note too the last paragraph of the description:

In case it's hard to quantify the impact in dollars, the following is a necessary but not sufficient condition for this market resolving YES: I ask friends/family who pay zero attention to AI developments and they are extremely aware of the thing.

OpenAI being worth a trillion dollars in 2023 (but without a large scale impact on people's lives in 2023) would almost certainly not reach the notice of the people who "pay zero attention to AI developments", let alone to the extent that they are "extremely aware" of it.

hassabis has called for ai to slow down. I doubt this will make that much of a difference on most metrics, but this one in particular is a lot less likely without their help.

update: okay starting to become confident that even a model capable of doing so won't

Related market that's specific to large language models:

Should be a lower bar since there doesn't need to be a trillion dollar+ impact to resolve positively, but it does require LLMs to be used.

Why is a technological singularity irrelevant for betting purposes? Shouldn't that make this resolve YES?

predicted YES

@IsaacKing can we get a clarification on how this resolves after a friendly singularity? I have a lot of plans for stuff to spend my m$ bragging rights on, gotta make sure I have as high an m$ score as possible on my full dive vr avatar in post-human extropia

How should I interpret "sudden" in this title? Is there any massive impact that won't count as sudden? (For example, if AI art just continues to gradually but rapidly improve until it reaches "superhuman" by the end of 2023, does that still count?)
predicted YES
@EigilRischel Good question, let's say yes, that would count! But it has to be so superhuman that it kinda turns the industry upside down. That's the "trillion dollar" part.
This is currently higher that https://manifold.markets/Forrest/will-ai-have-a-sudden-trillion-doll-8d2fc453ac75, which is clearly wrong.
predicted YES
Anyone have examples of past trillion-dollar events? Smartphones would be one, I think. Maybe it's a tall order to point to a single 1.5-year period when a trillion dollars of value was added. If so, that may make it less likely for this to resolve to YES. Picking resolution criteria for predictions sure is hard. But very roughly I'm trying to ask this: PaLM and DALL-E are scarily good. Will things like that result in obvious changes in normal people's everyday lives by the end of 2023?
predicted YES
Fair enough, we can add an implicit "conditional on lack of apocalypse" to this.
Unless it's true without causing human extinction, which seems vastly more likely?
@horse Maybe "way more likely" rather than "vastly"? There seems to be consensus among computer scientists that it's at least a risk worth taking seriously.
Mana is worth less if this is true due to human extinction.
© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules