
10GW AI training run before 2029?
Plus
14
Ṁ14272029
43%
chance
1D
1W
1M
ALL
In "Situational Awareness: The Decade Ahead", Leopold Aschenbrenner predicts that the largest AI training clusters will consume 10GW of electricity in ~2028.
This market resolves YES if a training run of a single AI model consumes 10GW+ of power sustained through most of the training run. This power cost includes overhead to run the data center, such as cooling.
This is one of a series of markets on claims made in Leopold Aschenbrenner's Situational Awareness report(s):
Other markets about Leopold's predictions:
This question is managed and resolved by Manifold.
Get
1,000and
3.00
Sort by:
Related questions
Related questions
AI: Will someone train a $10T model by 2100?
59% chance
1GW AI training run before 2027?
64% chance
100GW AI training run before 2031?
37% chance
$1T AI training cluster before 2031?
65% chance
$100B AI training cluster before 2029?
82% chance
Will any developed country establish a limit on compute for AI training by 2026?
21% chance
Will AI image generating models score >= 90% on Winoground by June 1, 2025?
82% chance
Limits on AI model size by 2026?
18% chance
Will an international pause on large AI training runs be in effect on Jan 1, 2028?
5% chance
AI: Will someone train a $1T model by 2080?
62% chance