OpenAI plans to create automated researcher by 2028: Altman

OpenAI expects its AI systems to reach the level of an intern research assistant by September 2026 and a fully automated AI researcher by 2028, CEO Sam Altman said during a livestream on October 28.

The company has completed its shift to a public benefit corporation, moving away from its original nonprofit model to enable greater capital raising and expanded research efforts.

Jakub Pachocki, OpenAI’s chief scientist, described the target system as one that can independently deliver large research projects.

Pachocki said it may be possible for deep learning systems to surpass human intelligence in many areas within a decade, though this remains a prediction.

OpenAI aims to achieve these milestones by improving algorithms and scaling up the computational resources used by its models.

The company’s nonprofit foundation will own 26% of the for-profit arm and manage research direction and safety initiatives, with a US$25 billion commitment to AI-driven disease research.

OpenAI plans to invest in 30 gigawatts of infrastructure, representing a US$1.4 trillion obligation in the coming years.

🔗 Source: TechCrunch

🧠 Food for thought

Implications, context, and why it matters.

The claimed $1.4 trillion obligation lacks evidence of secured funding or infrastructure commitments

OpenAI set a 30 gigawatts (GW) build plan, yet only 4.5 GW sits under a $30 billion Oracle deal 1. That is 15% of the goal. Reports do not show financing, power deals, or permits for the other 25.5 GW.
The Stargate initiative, an OpenAI linked data center build program, has raised about $50 billion from partners including Oracle and MGX (an Abu Dhabi backed AI infrastructure investment platform) 1. The gap raises timing doubts.
Power supply limits AI growth 2. Multi gigawatt builds need permits, grid hookups, and community meetings in many states 3. The 30 GW timeline stays uncertain without utility deals and regulatory approvals.

The $25 billion disease research pledge may open near-term chances for biotech and research groups

OpenAI cites a $25 billion commitment for AI disease research tied to its nonprofit foundation. That pool could fund grants and partnerships for groups in AI and biomedicine.
Biotech firms plus academic labs should watch for Request for Proposals (RFPs) and details. Early moves could land support like the $50 million People-First AI Fund, which offers up to 10% of operating budgets for eligible nonprofits 4.
Startups and university labs building AI for drug discovery, personalized care, or trial optimization should start drafting. For reference, the People-First AI Fund limited eligibility to U.S.-based nonprofits with annual budgets from $500,000 to $10 million 4.