Sam Altman: OpenAI expects to achieve AGI by 2026

On 28 August, according to Forbes, Chief Executive Officer Sam Altman of OpenAI, in an interview intime magazinetime, stated that the company expected to achieve its defined internal AGI system by 2026 — that is, AI that matches or exceeds human performance in most economic value tasks. The Chief Research Officer, Mark Chen, described the company as “80 per cent of its journey”, and the co-founder, Greg Brockman, thought that two years later, the period might be seen as the moment when AGI was born. The Astra model, to be launched by OpenAI, has been able to carry out the experiment and complete the work that previously required one week of human researcher work through internal automated AI research interns benchmark testing. Altman claims that Astra will be “the first model to invent new things in a meaningful way”. OpenAI is developing multi-purpose equipment (including desktop equipment introduced in early 2027), its own reasoned chip, Jalapeño, and human robotics, with the long-term goal of creating a universal AI subscription service for voluntary referrals and assignments。