OpenAI has just been marked 10 years With big news: GPT-5.2 releasedA model designed to master knowledge work, and a bold prediction from CEO Sam Altman that superintelligence is now practically inevitable over the next decade.
The new model, which follows an internal “Code Red” directive to accelerate development, introduces significant improvements over the previous model to make it better at executing complex, real-world tasks from start to finish.
This release signals that OpenAi is moving from abstract IQ scores to metrics that track how well AI works in the real world.
To discuss the potential impact of GPT-5.2 and OpenAI’s first impressive decade in business, I spoke with Paul Roetzer, founder and CEO of SmarterX and the Marketing AI Institute. Episode 186 of the Artificial Intelligence Show.
intelligence beyond reason
Initially, the AI industry measured progress using benchmarks that looked like standardized tests, essentially IQ tests for machines. But according to Roetzer, we’ve hit a ceiling with those metrics.
“Tests of intelligence are basically saturated,” Roetzer says. “When you’re trying to evaluate these models against standardized tests that humans take, AI already exists.
This is basically at top human level, if not beyond top human level in many of these tasks. So when we are talking about increase in IQ points it is really hard for all of us to feel the difference.
Here comes GPT-5.2, which adds value to GDP
With the release of GPT-5.2, OpenAI is leaning heavily toward a new benchmark that evaluates the model’s ability to perform 1,300 specific tasks from 44 businesses that contribute to GDP. The result is real-world deliverables, such as creating legal briefs, engineering blueprints and nursing plans.
Result Eye-opening: OpenAI reports that GPT-5.2 Thinking achieved a win rate of about 71% against human experts in a direct comparison.
“We need to start measuring based on actual work because that’s how we know when economic disruption is coming,” Roetzer says. “And I would say we are there.”
Faster, cheaper, “better” than humans?
The implications of this GDPVal, as the benchmark is called, go beyond quality. Data from OpenAI shows that Frontier models can complete these professional tasks nearly 100 times faster and 100 times cheaper than human experts in these fields.
In its announcement, OpenAI described it as a tool to “unlock even more economic value for people”, emphasizing how the model can help with spreadsheets, presentations, and coding.
“Note that they don’t say we want to change more jobs,” Roetzer says.
However, the subtext related to human job displacement is hard to ignore.
a changed mission
OpenAI released GPT-5.2 during its 10th anniversary in a look back at the company’s origins.
Founded as a non-profit in December 2015, its original mission was to advance digital intelligence to benefit humanity, without the need for profit. The original announcement vowed to distribute electricity equally and share patents with the world.
Compare that vision with now, and the contrast is obvious.
Referring to the Founding Statement, Roetzer says, “There is nothing true in that paragraph anymore.”
Today, OpenAI is a major commercial force, competing against global competitors including Google to achieve Artificial General Intelligence (AGI). one in memorial postSam Altman, reflecting on the journey, said that the company is now “almost certain to create superintelligence in the next 10 years.”
OpenAI’s evolution from an idealistic non-profit research lab to a powerhouse driving the global AI arms race is one of the most important business stories not just of the last decade, but of all time.
