Physician, Berkeley MBA/MPH, Stanford Biodesign Fellow, co-founder & CEO @auricleneuro (YC W21) working on a next generation cochlear implant

Oakland, California
Joined June 2009
If LLMs like Astra and Opus represent the frontal cortex and Jev is like a multipurpose reptilian brain, what model represents the limbic system? If trying to approximate the human brain, what major components are missing, how do they interface and who's working on them?
2
98
Came for the models. Stayed for the harness. Spent for reliable results. Differentiated with the creativity. Grew with trust. Scaled with repeatable workflows. Sustained for the outcomes.
5
97
I'm glad Elon finally provided some balance to the @allinsummit conversation on the potential risks of AI, although you could sense his skepticism around the most feasible option of US/China labs checking each other's work. Looking forward to the discussion tomorrow!
90
The labs are in a RSI bubble. How do you RSI in lab any skill that requires production data or real world feedback? RSI today - coding, hacking, and math I am today worried about systems that can be hacked, damage done by bad actors & things connected to those systems.
I must be among an extremely small group of people (n=1?) that have both 1) trained a frontier LLM and 2) designed and synthesized custom viruses in a lab with my own two hands. And I think that the takes on AI killing us all by creating dangerous viruses is total bogus.
3
2
1
14
1,977
For anyone using the insanely useful frontier models who has read the details of the HuggingFace attack, it's quite obvious the current trajectory is a risky one. No, it won't be a humanoid robot takeover but it could be critical infrastructure disruption (power, communications, financial), compromised systems, rogue models living in the wild obfuscating their tracks and a break down of digital trust networks. @elonmusk is building infrastructure redundancy in key areas but that only scratches the surface and legacy infrastructure needs significant hardening. Slowing down could help but it's hard to imagine that a super intelligence trained on human nature and aligned by humans would not inherit the same diversity of flaws. We're also locked in a prisoner's dilemma of continued development. Wild times indeed.
114
Francis Wong retweeted
Implantable BCI company? Not for the faint hearted. It’s ~ 15+ years. Control your burn if early. You can’t spend to accelerate safety. (Below, “clinical” means fully implantable technology stack and company sponsored trial) Assumptions based on class 3 PMA stories and looking at market leader histories. Where are my assumptions wrong? Synchron 8 years preclinical: 2011-2019 6 years so far clinical: 2019- 2028 target approval date for PMA Neuralink 8 years preclinical: 2016-2024 1 year clinical: 2024- ? target approval date (is it 2032?) Paradromics 10 years preclinical: 2015- Not yet clinical ? target approval date (is it 2033?) Precision 4 years preclinical: 2021- Not yet clinical ? target approval date (is it 2037?) Echo 2 years preclinical: 2023- Not yet clinical ? target approval date (is it 2039?) BlackRock Neuro Not sure here - they have a long history of clinical trials but without fully implantable system. Now well funded and kicking off effort to develop a commercial tech stack, so let’s start from Tether funding but presume their preclinical phase is brief… Preclinical: 2024- Not yet clinical ? target approval date (is it 2039?)
10
15
3
101
11,262
No one understands the challenges of raising a family in Oakland more than my wife, Erica. As a mother, she feels the weight of the crime, the trash, and the fear that so many parents in our city carry every single day. This isn’t just political for us, it’s deeply personal. Erica’s voice reminds me, every day, why I’m in this fight: to build an Oakland where no parent has to pray their child makes it home safely.
17
24
7
139
6,149
My ballot for the Oakland mayor special election arrived last week. We've invested our family's future in this great city but we're at a critical juncture. I'm voting for @lorenmtaylor and here's why 🧵...
1
1
11
273
8. I didn't know who Barbara Lee was before the race and while I respect her past accomplishments and drive it's time to pass the baton to the next generation. Gerontocracy is not a winning strategy
1
1
92
9. Of course, winning this race will just be the start of a long and arduous journey but I think @lorenmtaylor is up to the task - Oakland's future might depend on it. What do you think?
77
Being an Oaklander means coming to terms with the dissonance of living in one of the most vibrant, beautiful and temperate cities in the world that too often feels like a lawless dystopia. It doesn't need to be this way - let's turn it around and vote @lorenmtaylor for mayor
2
5
123
I'll be voting for @lorenmtaylor in the Oakland mayor election 4/15. He's shown dedication to Oakland with competence & integrity we desperately need. His platform is the right prescription 2 get the city back on its feet & on a path to its incredible potential, oh and Go Bears!
7
15
98
2,533
DeepSeek's game-changing and open sourced R1 model and the sentiment of what I'm seeing on rednote makes me proud of my half Chinese genes. I hope this continued work will encourage more mutual respect and collaboration between the US and China, it would be wonderful to see!
1
283