AI Healthcare "Three-Kingdom Battle": Google DeepMind vs OpenAI vs China, Who Will Cure Cancer First?

Silicon Valley — If asked what AI's most valuable application scenario is, healthcare definitely ranks in the top three.

And on this track, three main players are competing intensely: Google DeepMind, OpenAI, and Chinese AI healthcare companies.

DeepMind: After Proteins

DeepMind's breakthroughs in AI healthcare are evident to all.

In 2024, DeepMind's AlphaFold successfully predicted structures of almost all known proteins — a major breakthrough in biology for 50 years.

Now, DeepMind is setting sights on cancer.

"We're training AI to predict cancer cell behavior patterns," DeepMind's lead said at a recent academic conference. "The goal is to predict cancer metastasis and recurrence in advance."

What does this mean? Doctors can formulate more precise treatment plans at early cancer stages.

OpenAI: Healthcare Assistant

Different from DeepMind, OpenAI's strategy is "healthcare assistant."

OpenAI cooperated with multiple hospitals to develop GPT-based healthcare assistants. These AI assistants can help doctors:

Quickly read medical records, extract key information

Assist diagnosis, provide similar case references

Generate treatment suggestions, reduce doctor workload

Quickly read medical records, extract key information

Assist diagnosis, provide similar case references

Generate treatment suggestions, reduce doctor workload

"We're not replacing doctors," OpenAI's healthcare lead said. "We're making doctors' work more efficient."

China: Application Landing

Chinese AI healthcare companies' strategy leans more toward application landing.

From AI medical imaging to intelligent consultation, from drug discovery to hospital management — Chinese AI healthcare applications are rapidly spreading.

"China's advantage is large data volume, many application scenarios," an investor said. "We can iterate quickly, verify quickly."

Data shows China's AI healthcare market grew 80% in the past year, expected to exceed 100 billion RMB by 2027.

Focus of "Three-Kingdom Battle"

Where do the three players compete?

Data: Who has more high-quality medical data has better models

Compute: Training large models requires massive compute

Application: Who can land AI products in medical institutions faster

Data: Who has more high-quality medical data has better models

Compute: Training large models requires massive compute

Application: Who can land AI products in medical institutions faster

Who Has More Advantage?

DeepMind: Has technical accumulation, but leans toward basic research OpenAI: Has model advantage, but relatively new in healthcare China: Fast application landing, but gap in basic research

"I think there will be major breakthroughs within three years," a healthcare AI investor predicted. "But specifically which team, hard to say."

Epilogue

At a medical conference, I met an oncologist with 30 years of experience. He told me he's now used to AI-assisted diagnosis.

"AI helped me detect dozens of early lesion cases," he said. "Before, it was easy to miss with the naked eye."

Perhaps this is the meaning of AI healthcare: not replacing doctors, but making good doctors even better.

Reference: Nature Medicine, TechCrunch, Wall Street Journal