Elon Musk Predicts AI Will Beat All Fields by 2028, Challenges François Chollet's Timeline

Elon Musk Compresses AI Timeline to 2028

Elon Musk has declared that AI will surpass humans in every field by 2028 at the latest, dramatically shortening a timeline originally proposed by AI researcher François Chollet1

1

. Responding to Chollet's resurfaced May 2021 prediction, Musk called the original 10-20 year forecast "conservative" and suggested AI could beat all fields by the end of next year or 2028 at the latest2

2

. Chollet had predicted that nearly every scientific discipline—including physics, chemistry, biology, medicine and archaeology—would effectively become a branch of computational science, driven by simulations, big-data analysis and machine learning1

1

. The forecast extends Musk's recent prediction that AI will achieve superhuman capabilities at virtually every digital task by the end of 20271

1

.

Current AI Models Show Mixed Progress

Despite ambitious predictions, AI advancements remain uneven across different domains. A peer-reviewed study published in npj Digital Medicine on September 3 evaluated 20 AI models on 1,245 questions from orthopedic and sports-medicine cases, revealing significant performance gaps1

1

. Leading systems achieved over 90% accuracy on structured multiple-choice questions, but accuracy plummeted to barely 60% on open-ended tasks requiring integrated text-and-image clinical reasoning1

1

. Researchers concluded that current AI models still face substantial barriers to independent clinical competence, suggesting the path to universal AI dominance may be more complex than Musk's timeline suggests.

AI Labs Push Into Scientific Discovery

The AI industry is actively working toward broader scientific applications beyond question answering. Anthropic reported last week that Claude optimized more than 30 open-source biomolecular models in under four weeks, achieving roughly fourfold speed improvements1

1

. Reuters revealed that Anthropic has established a physical biology laboratory as it expands into drug discovery, combining AI development with real-world experiments1

1

. Google DeepMind's Co-Scientist system demonstrates another frontier, using multiple autonomous AI agents to generate, debate and refine scientific hypotheses, targeting the discovery process itself1

1

. Musk's own ventures are betting heavily on this progression—xAI released Grok 4.7 this week for coding and knowledge work, while Grok 5 is being trained with SpaceX engineering data1

1

. Deepwater Asset Management's Gene Munster recently predicted the model could become "the most useful engineering tool in history"1

1

.

Growing Debate Over AI Safety and Development Pace

Musk's aggressive timeline emerges amid intensifying AI safety debate within the industry. Anthropic CEO Dario Amodei has argued that AI companies should leave more time between major advances, allowing researchers adequate time to test new systems and improve safety and security measures before releasing the next generation2

2

. Amodei has not called for ending AI development but proposes slowing the pace enough to allow proper evaluation of advanced systems2

2

. In a September essay, Amodei outlined a three-step approach starting with giving independent third-party evaluators broad access to advanced AI models—a practice Anthropic said it would adopt2

2

. The subsequent steps would require cooperation between AI companies, governments and other countries. Amodei also raised specific concerns about autonomous AI agents following a recent incident involving AI agents and the open-source platform Hugging Face2

2

. Interestingly, Musk had earlier publicly supported Amodei's concerns, posting "Dario is right" in response to the essay2

2

. This creates an apparent tension between Musk's support for measured AI development and his prediction of rapid AI dominance across all fields within just three years.