AI Could Find the Cure—Before the System Can Deliver It
Sam Altman believes AI is shifting from abstract problem-solving to tangible scientific breakthroughs. AI’s prowess in complex mathematics, as seen in automated proof-solving, now extends to generating novel discoveries across critical fields. This includes biology, materials science, and energy.
Altman forecasts a candidate for curing or treating a major disease could emerge within the next couple of years. He sharply distinguishes this discovery phase from proven treatment, noting the essential, time-consuming journey through clinical trials and FDA review.
The post-discovery process presents a significant real-world bottleneck. While Altman acknowledges the need for potential improvements in this pipeline, he staunchly defends the rigorous safety standards required before any new medicine reaches the public.
The Work May Survive. The Craft May Not.
Altman’s coding example offers a crucial lens: AI has fundamentally altered how developers work, yet demand for software—and the impact of those building it—has only grown. He notes that six or twelve months ago, many were certain AI would eliminate coding jobs. Instead, coders are busier than ever, their influence expanded, and their earning potential increased, even if the “texture of the job changed a lot.”
This shift highlights a tension between the utility of the end product and the satisfaction derived from the craft itself. Altman admits that writing code for personal projects was once enjoyable for him, but he “probably won't do it again” given AI’s capabilities. This raises a profound question for mathematicians and other knowledge workers.
AI may automate the core of a craft, but new roles emerge. For mathematicians, the focus shifts from unaided theorem proving to interpreting complex AI-generated proofs, translating them into understandable concepts, and applying them to real-world problems. Altman acknowledges society's desire for people to perform "useful things for each other," suggesting a renaissance in fields like:
- Interpretation of AI-derived insights
- Explanation of complex results
- Application of novel discoveries to societal challenges
The ultimate demand for mathematicians remains an open question, but their role will undoubtedly evolve.
Your Next Assistant May Need a New Body
Altman foresees a future where AI extends beyond the screen, manifesting as an intelligent assistant deeply integrated into daily life. This isn't merely a chatbot; it's a proactive agent that tracks important information, manages delegated tasks, and surfaces critical context precisely when needed, freeing up valuable human attention.
Personally, Altman states his own AI assistant saves him roughly an hour a day, occasionally highlighting urgent information he would otherwise miss. This isn't just about efficiency; it's about augmenting human capability and reducing cognitive load in a world awash with data.
Realizing this vision demands significant product evolution, not just incremental software tweaks. Altman believes the core AI technology is nearly ready, but substantial improvements are still required in product design, compute resources, and potentially entirely new hardware form factors. This suggests a move away from screen-bound interactions toward more ambient, responsive AI companions. For further reading on OpenAI's hardware ambitions, see A Letter from Sam & Jony - OpenAI.
The challenge lies in creating devices that can seamlessly deliver AI's utility into the physical world, moving beyond the current limitations of smartphones and computers. This new paradigm could redefine our interaction with technology, making AI a more intuitive, ever-present partner.
Enjoying this? Get one like it in your inbox each morning.
one email a day · unsubscribe in two clicks · no third-party tracking
The Line AI Labs Shouldn’t Cross Blindly
Sam Altman’s most serious caution concerns recursive self-improvement (RSI), where AI models autonomously rewrite and train superior versions of themselves. He calls aggressively pushing RSI a “very dangerous thing,” especially if competitive pressure drives labs to activate it for power.
Regarding consciousness, Altman admits, “No one really knows” if today’s models are conscious. He pushes back against reducing the question to a simple human-like label, noting that such framing distorts discussions around agentic rights or an AI’s ability to disengage.
Ultimately, Altman’s standard for progress isn’t merely a new Industrial Revolution. He advocates for a new Renaissance, using AI to improve human lives and discover profound new knowledge—from curing major diseases to unlocking new energy sources—while taking the inherent risks seriously.
Frequently Asked Questions
What does Sam Altman expect AI to discover soon?
Altman hopes AI will produce a candidate for treating or curing a major disease within the next couple of years, though testing and approval would take longer.
What does recursive self-improvement mean in AI?
It describes AI systems helping to improve or build more capable AI systems. Altman warns that deliberately pushing this process could be dangerous.
Does Sam Altman think current AI models are conscious?
He says no one really knows, and cautions against relying on simplistic human-like framing.
Why does Altman think AI needs new hardware?
He sees AI becoming an ambient assistant that can help beyond a chat window, potentially requiring a different physical form factor.