OpenAI is now ten years old
On the same day it launched its most powerful GPT to date, OpenAI boss Sam Altman reflected on a decade of research and outlined its path toward AGI.
OpenAI has marked its ten-year anniversary and outlined progress toward artificial general intelligence in a company blog post.
Sam Altman wrote that the organisation began with a small, optimistic team and a focus on learning by doing. He noted early culture and technical work aimed at safety and robustness and highlighted several 2017 results, including Dota 1v1, the unsupervised sentiment neuron, and reinforcement learning from human preferences.
"OpenAI has achieved more than I dared to dream possible; we set out to do something crazy, unlikely, and unprecedented," Sam Altman wrote.
He added that the group launched ChatGPT three years ago and later released GPT‑4, developments that have accelerated public use and regulatory and operational responsibilities.
Altman said the company adopted a strategy of iterative deployment to let society and the technology co‑evolve.
"In ten more years, I believe we are almost certain to build superintelligence," he wrote, and he described the recent period as intense and requiring difficult execution.
The post said OpenAI is grateful to users and customers for adopting early versions of its technology and that the company expects further advances over the next couple of years while continuing to pursue its mission to ensure AGI benefits all of humanity.
The Recap
- OpenAI reflected on ten years and progress toward AGI.
- Launched ChatGPT three years earlier and later released GPT‑4.
- Sam Altman said superintelligence is likely within ten more years.
