the story i roughly understand is that this was within Epoch’s mandate in the first place because they wanted to forecast on benchmarks but didn’t think existing benchmarks were compelling or good enough so had to take matters into their own hands. Is that roughly consensus, or true? Why is frontiermath a safety project? i haven’t seen adequate discussion on this.
They say it was an advanced math benchmark to test the limits of AI, not a safety project. But a number of people who contributed would have been safety-aligned and would not have wanted to if they knew OpenAI will have exclusive access.
the story i roughly understand is that this was within Epoch’s mandate in the first place because they wanted to forecast on benchmarks but didn’t think existing benchmarks were compelling or good enough so had to take matters into their own hands. Is that roughly consensus, or true? Why is frontiermath a safety project? i haven’t seen adequate discussion on this.
They say it was an advanced math benchmark to test the limits of AI, not a safety project. But a number of people who contributed would have been safety-aligned and would not have wanted to if they knew OpenAI will have exclusive access.