Chamath Palihapitiya
· Chamath Palihapitiya
· January 31, 2026 at 17:01
· ⏱ 5 min read
| Read on Substack ↗
Summary
=== SUMMARY ===
•Foundational AI models are rapidly expanding beyond the digital realm (language, images) into the physical (robotics) and biological (genomics) domains, creating significant, sector-wide investment catalysts.
•Progress by private technology leaders like Neuralink and Figure AI serves to de-risk and accelerate entire emerging industries (Brain-Computer Interfaces, Humanoid Robotics), creating opportunities in public companies within those ecosystems and their supply chains.
•The AI arms race is a global phenomenon, with intense competition in both the US (Google, Microsoft) and China (Alibaba, ByteDance) driving a secular growth trend for enabling technologies, particularly high-performance semiconductors.
Summary
The newsletter highlights three major tech breakthroughs: Neuralink's rapid trial expansion and new BCI products (Telepathy, VOICE, Blindsight), Figure AI's Helix 02 achieving full-body autonomy in humanoid robots, and Google DeepMind's AlphaGenome model for DNA analysis outperforming existing tools. These developments signal accelerating convergence of AI, neuroscience, and robotics, with Alphabet (via DeepMind) positioned to lead in genomic AI research, while private companies Neuralink and Figure AI push hardware frontiers.
•Neuralink expanded human clinical trials to 21 participants (up from 12), with signal quality improved in 18 of 20 recent cases.
•Neuralink opened a new VOICE trial aiming to restore speech at 140 words/min and Blindsight received FDA Breakthrough Device designation.
•Figure AI's Helix 02 model replaces 100,000 lines of C++ code with a neural network trained on 1,000+ hours of human motion, enabling full-body loco-manipulation.
•Helix 02 uses Palm Cameras and tactile sensors (sensitive to 3 grams) to handle delicate tasks like picking a pill from a bottle.
•Google DeepMind's AlphaGenome outperformed existing models in 24 of 26 evaluations, analyzing up to 1 million base pairs to predict DNA variant effects.
•AlphaGenome is released free for non-commercial use to accelerate disease understanding and personalized gene therapies.
The article reports that Google DeepMind's AlphaGenome outperforms existing models in 24 of 26 evaluations and is released for free to accelerate genomic research, underscoring Google's AI leadership
The article reports that Google DeepMind's AlphaGenome outperforms existing models in 24 of 26 evaluations and is released for free to accelerate genomic research, underscoring Google's AI leadership in a new scientific domain beyond proteins.
Risk: AlphaGenome is a research tool with no near-term revenue; commercial application timeline is uncertain and regulatory hurdles for gene therapies remain high.
This newsletter, published January 31, 2026,
features Chamath Palihapitiya
discussing GOOGL.
1 trade idea extracted by AI with direction and confidence scoring.