About me

I am Patrick J. Mineault, PhD, neuroscientist and AI researcher.

I research how the brain processes visual information through advances in AI. By building machines that can see the world the same way we do, I believe we’ll be able to design new interventions to help people with neurological disorders.

I did a PhD a visual neuroscience at McGill with Dr. Chris Pack and a postdoc at UCLA with Dr. Dario Ringach. I studied how the brain creates meaningful visual representations which can drive behaviour. I worked at Google, studying at how people view and interact with webpages. I was also a Brain Computer Interface engineer at Facebook Reality Labs, building a BCI that allows you to type with your brain.

I took a sabbatical to help build the first edition of Neuromatch Academy as CTO. It’s an online summer school in computational neuroscience and AI. The first edition brought together 191 TAs and 1700 students from 60 countries learning full-time for 3 weeks.

I co-founded a startup, Blindsight Therapeutics, where I was CTO and built the IP portfolio. Our aim was to help people with cortical blindness through advances in neuro AI.

I’m always happy to take on new, manifestly important projects that are nearly impossible. I am also available as a speaker. My email is patrick.mineault@gmail.com.

Work

2020-2022 – CTO and co-founder, Blindsight Therapeutics, Montreal

Building a treatment for people suffering with cortical blindness, inspired by advances in neuro AI.

2017-2019 – Brain Computer Interface Engineer, Facebook, Menlo Park

Building a brain-computer interface that will allow people to type with their thought. Some press coverage: [1] [2] [3]

2015-2017 – Software Engineer and data scientist, Google, Mountain View

Helping organize the world’s information and make it universally accessible and useful. Ran some of the largest psychophysics study ever.

2014-2015 – Postdoctoral researcher at the Dario Ringach lab, David Geffen School of Medicine, UCLA.

Studied representation in the visual cortex of mice, looking at how stimuli are differentially encoded during locomotion.

2008-2014 – Doctoral researcher at the Pack Lab, McGill University, Montreal.

I studied visual representation and decision making at the level of single neurons, populations of neurons, and psychophysical observers [PhD thesis].

Education

  • PhD (2014), Integrated Program in Neuroscience, McGill University. Thesis: Parametric modeling of visual cortex at multiple scales. [PDF]
  • B.Sc., Physics and Mathematics (2007). McGill University

Papers

[Always up to date list on Google Scholar]

Patents

  • Mineault, P.J. Neural decoding with co-learning for brain computer interfaces. US11314329B1. Assigned to Meta Platforms Inc.
  • Chevillet, M.A., Mineault, P.J., Mugler, E.M., Jennings S.Y.K. Wearable brain computer interface. US11301044B1. Assigned to Meta Platforms Inc.
  • Chevillet, M.A., Dugan, R.E., Jennings S.Y.K., Choma M.A., Tiecke T.G., Mineault, P.J., Mugler, E.M., Wettersten, H. Brain computer interface for text predictions. US10795440B1. Assigned to Meta Platforms Inc.

Speaking experience

Elsewhere on the web

Twitter: @patrickmineault
LinkedIn
Neurotree
Github

About this blog

I have been writing about neuroscience, AI, programming and data science on this blog since 2008.

Licensing

All of the code snippets on this blog are authored by me – unless otherwise indicated – and are licensed under an MIT license, which means you can use them without asking me about it. All the text is CC-BY-2.0, which means you can include it in a textbook, a paper, or repost it elsewhere and potentially modify without asking permission as long as you maintain an attribution line.

2 comments

Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s