半吊子区块链学屌. 清华哲学, 奥大神学, 维大经济政策. Blockchain, AI, technology optimist. Philosophy. Economics. Psychology. Policy. - (might ackchually just be a dog on the internet)
It would be interesting to compare AGI capability tests with nonverbal IQ tests, such as the RAPM, and see how current definitions of AGI overlap with g (general intelligence factor). No test spamming, maximum two attempts per model, just like Mensa exams. #AGI#AI#o3#GPT
Today OpenAI announced o3, its next-gen reasoning model. We've worked with OpenAI to test it on ARC-AGI, and we believe it represents a significant breakthrough in getting AI to adapt to novel tasks.
It scores 75.7% on the semi-private eval in low-compute mode (for $20 per task in compute ) and 87.5% in high-compute mode (thousands of $ per task). It's very expensive, but it's not just brute -- these capabilities are new territory and they demand serious scientific attention.
@littmath Diverse ways of learning should be encouraged as long as the student arrives at a correct understanding. As a child, my proof that the hypotenuse was the longest side of a right triangle was to rotate a circle on both ends of it, unfortunately, the teacher dismissed my solution.
Australia is proudly home to 1.4 million people of Chinese ancestry.
They're an important part of our national story, and an integral part of our national identity.
Emerging fields are often ripe with opportunities for mediocre minds to exploit fringe political groups through adversarial propaganda, as a lack of public understanding is necessary to reduce complex ideas into marketable rhetoric designed to purposely ignore the bigger picture.