Will we throw away (almost all) convolution layers in 2024?
16
90
Ṁ358Ṁ1k
2025
17%
chance
1D
1W
1M
ALL
Will we throw away convolution layers in computer vision this year?
We have seen DiT (Sora) replacing the U-Net with transformers.
I predict we will see more this year until all convolution layers are replaced with something that does not explicitly tell models to look locally.
I will consult public and expert opinions on this matter at EOY 2024.
Get Ṁ600 play money
More related questions
Related questions
By EOY 2026, will it seem as if deep learning hit a wall by EOY 2025?
28% chance
Will there be any large-scale protests about content generated by Deep Learning by the end of 2025?
40% chance
Will OpenAI report that it's no longer heavily GPU limited in 2024?
27% chance
Will it be possible to disentangle most of the features learned by a model comparable to GPT-3 this decade? (1k subsidy)
55% chance
Will transformers still be the dominant DL architecture in 2026?
57% chance
Will it be possible to disentangle most of the features learned by a model comparable to GPT-4 this decade?
39% chance
Will any image-generation AI be able to consistently draw simple polygons by the end of 2024?
73% chance
Will Huggingface a fold by 2028?
42% chance
Will AI pass the Winograd schema challenge by the end of 2024?
78% chance
Will a new type of AI model replace transformers as the state-of-the-art artificial intelligence technology by 2024?
26% chance