Will a large language model beat a super grandmaster playing chess by 2028?
By the end of 2026, will we have transparency into any useful internal pattern within a Large Language Model whose semantics would have been unfamiliar to AI and cognitive science in 2006?
Will a 10B parameter multimodal RL model be trained by Deepmind in the next 12 months?
HumanEval 90% #1: Will pass@1 performance on the HumanEval benchmark be >= 90% by 2024?
Will there be a Forward-Forward Algorithm based neural network with >65% Top 1 Accuracy on Papers With Code's ImageNet leaderboard by 2024?
Will reinforcement learning overtake LMs on math before 2028?
Will a model costing >$30M be intentionally trained to be more mechanistically interpretable by end of 2027? (see desc)
When will tinygrad train its first MLPerf qualifying model?
🐕 Will AI Achieve Significantly More, "Embodiment" by end of 2023?
Will we learn by EOY 2024 that large AI labs use something like activation addition on their best models?
By 2025 end, a model exhibits action recognition (video) equivalent to human level accuracy on Something Something V2?
Will language models be able to solve simple graphical mazes by the end of 2025?
Will a news article containing the string "love in the time of large language models" be published in 2023
Short Term AI 2.3: By January 2024, will SOTA on MATH minival be >= 85%?
Will Transformer based architectures still be SOTA for language modelling by 2026?
Will more than 20 organizations publicly train large language models by 2024?
By 2030, will large language models still be at the peak of AI? [DRAFT]
Will a new best accuracy for ImageNet classification be achieved before the end of 2023?
Will Metaculus have built-in support for reflective latent variables by 2025?
Will anyone train a 50B parameter+ RetNet by the end of 2023?