🤖 Which 5 AI advancements in 2024 will be the most important? [Free response]
Basic
71
5.3k
2025
66%
Gemini 2.0
63%
OpenAI's GPT 5
49%
OpenAI's Sora
34%
OpenAI's GPT 4.5
34%
Google's Video Prism
33%
Google's Gemini 1.5
22%
Anthropic's Claude 4
20%
Midjourney video
17%
Tesla FSD V13 (next version)
16%
Google's Gemma
16%
Meta's Llama 3
16%
Anthropic's Claude 3
15%
OpenAI's Q*
15%
Apple's on-device model
15%
Google's Lumiere
14%
Deepmind's AlphaGeometry
13%
Mistral's mistral-next
12%
John Carmack's Keen Technologies AGI
11%
Google's Gemini 1.0 Ultra
7%
Nvidia's RTX 5000 series

(This market is in beta, so the description will likely change over the next few days as I figure out edge cases)

Free response! If you add an option, it must be limited to a single (or closely related family of) techniques, papers, models, or devices. Broad answers like "AI alignment", “a new attention model”, or answers that overlap with existing ones will be removed.

Advancements must have been announced, published, or released in 2024. So Gemini Pro doesn't count (released last December), but Gemini Ultra does. An advancement that was announced in 2024, but not released to the public would still qualify.

The top 5 results on a poll asking "Which AI advancements from 2024 were the most important" will resolve YES. The poll will remain open for 2 weeks. Options that have consistently traded at a low percentage on this market may be omitted from the poll. Options will be ordered by their ranking on this market.

Speculative or anticipated advancements may be added, as long as they are specific products or models made by a specific company. But if they are not announced, published, or released by the end of 2024, they will resolve NO and be left off the poll.

I won't trade on this market

Get Ṁ600 play money
Sort by:

With the gpt-4o announcement, what's a good formulation of "Her-like conversational ai"? Does it need to specify OpenAI/Google? (I think both have announced something, although Google wasn't set to release in 2025)

These answers are a bit too broad, since there can be multiple breakthroughs in one year, and I’d like each answer to be a specific advancement. Also, speculative answers have to be a specific product or model.

To make this one more consistent, I’m going to rename to “DeepMind’s earthquake prediction model”.