Social media influencers have kicked off a fierce debate over OpenAI’s new o3 reasoning model, with some of them raising concerns about its high cost and the potential for overhyping its artificial general intelligence (AGI) capabilities.
Several of them also expressed awe at its impressive performance across various benchmarks, including exceeding human performance on coding and scientific reasoning tasks, according to GlobalData, a leading data and analytics company.
Influencers are particularly captivated by o3’s ability to perform complex tasks autonomously, potentially exceeding human capabilities in nearly every cognitive domain, and its demonstration of a ‘private chain of thought’ approach, allowing for more reliable answers, said Shreyasee Majumder, social media analyst at GlobalData.
However, there is significant debate surrounding the model’s cost-effectiveness and its potential for generalizability across different problem domains.
Some influencers highlight the massive compute required for o3, pointing out that while it may be a significant step forward, its real-world applicability is still limited due to its high cost.
“Additionally, some influencers challenge the claims of achieving ‘AGI’, arguing that while o3 demonstrates impressive capabilities, it is not yet capable of true general intelligence and lacks robustness across all problem types,” said Majumder.
According to Elvis S, Cofounder and CEO at DAIR.AI, the hype around o3 is out of control.
“It’s not AGI, it’s not the singularity, and you definitely don’t have to change your worldview. In fact, the public doesn’t even have access to the models so how can anyone claim any of the above. I appreciate how the OpenAI researchers presented o3,” said Elvis.
Aaron Levie, CEO at Box, added that OpenAI’s o3 model appears to be better at reasoning than any other model out there.
“It costs way more to operate, but that’s irrelevant. What is expensive today is cheap tomorrow. Quality is all that matters because you know that costs will always drop,” he mentioned.
Chamath Palihapitiya, CEO at Social Capital, said the AI model sets new performance records across multiple benchmarks, achieving 96.7 per cent on the American Invitational Mathematics Exam, 87.7 per cent on graduate-level science questions, and a Codeforces rating above the 99th percentile.
