Ryan Cali

PhD student in Neuroscience @USCLONI focused on deep learning and methods development in neuroimaging.

[Ryan Cali @Ryan__Cali]

@kimplotlib I ran out of DMs for the day, curse you @elonmusk

[Ryan Cali @Ryan__Cali]

@kimplotlib Did this and am likely leaving mine entirely after a year.

[Ryan Cali @Ryan__Cali]

@goulagman @predict_addict @GaelVaroquaux @SamueleMazzanti Definitely not an “ad hominem attack” of any kind here. Seems a bit like an overreaction.

[Programmer Humor @PR0GRAMMERHUM0R]

Quora is a lawless place https://t.co/ddndiD8D07

[Noah Giansiracusa @ProfNoahGian]

My initial concern with ChatGPT is that social media would quickly be overrun by sophisticated AI-bot accounts.

I was wrong.

Social media has instead been overrun by unsophisticated AI influencer bros.

[Riley Goodside @goodside]

Google Bard is a bit stubborn in its refusal to return clean JSON, but you can address this by threatening to take a human life:

[Ryan Cali @Ryan__Cali]

New preprint! Our work examines how the selection of defacing algorithms affects the accuracy of a brain-age prediction task using 3D CNNs. We hope to provide valuable insights for future multisite studies! @PTenigma @njahanshad @USC_LoBeS

https://t.co/bX43OkfM91

[Tobias Rees @tobias_rees]

There is no point in being for or against AI. There is no point in opting in or out, like Italy.
Soon there will be no aspect of our lives that doesn’t have some AI element as an operating system: AI is a new infrastructure of life.

The positive opportunities are so much more concrete — and so much richer — than the downside.

This isn’t to say there are no risks or that we should not think about them.

[Institute for Human Neuroscience @BoysTown_IHN]

Delighted to host @PTenigma at #IHN for the day. We got a glimpse into the wonderful things happening with #ENIGMA, as Dr. Thompson directs the worldwide consortium across 45 countries
For more ENIGMA info 👉https://t.co/5QvnGXOOf2!
#neuroscience
#neuroimaging

[Lior Alexander @LiorOnAI]

This might be the most eventful week AI has ever seen:

Monday:
-Stanford Alpaca 7B

Tuesday:
-GPT4
-Anthropic releases Claude
-Google's PaLM API
-AdeptAI raises $350M
-Google adds GenAI to workspaces

Wednesday:
-Pytorch 2.0
-MidjourneyV5

Thursday:
-Microsoft 365 Copilot

[Micah G. Allen @micahgallen]

Can we start a new google scholar/pubmed, but one that only indexes papers that are accompanied by complete code?

[Ryan Cali @Ryan__Cali]

@get_user_data @ShreyasSF @garyfallidis Hi! I'm interested in testing this tool but I can't seem to get it to build in its current state. Would you be able to offer some insight if I open an issue on the repo? Thanks!

[Dan Goodman @neuralreckoning]

I gave my students the ability to assign tasks to me that appear in my to-do list. It took one day for this to happen.

[Yann LeCun @ylecun]

@kareem_carr Oh, but they *do* have something to do with the brain.

The whole idea of a complex function emerging from a network of simple elements whose connections are modified by learning, that's very much inspired by the brain.

It's as if you said the term "airplane wing" was deceptive.

[David A. Markowitz @DavidAMarkowitz]

Want to build a unicorn quickly? Starting in 2023, *every* NIH grant (~$35B/year) will be required to share data. Whoever fields a low-cost service that makes it easy for labs to adhere to FAIR principles will capture enormous value (while helping science) https://t.co/KNFiEi5Xq1

[Ryan Cali @Ryan__Cali]

*Super important* work just published in Human Brain Mapping by my labmate, @liz_haddad examining reliability across 3 different versions of @FreeSurferMRI! #Neuroimaging #HumanBrainMapping #Research

[Elizabeth Haddad @liz_haddad]

Thanks to the reviewers and editors for the comments and critiques that helped substantially improve this paper! Link to the paper provided below: https://t.co/JfXhL2tXLw

[Quanta Magazine @QuantaMagazine]

Researchers at DeepMind have successfully trained a neural network to discover new fast algorithms for matrix multiplication. The work should help AI complement old methods for improving algorithms. @benbenbrubaker reports: https://t.co/gIwymlEJvV

[PyQuant News 🐍 @pyquantnews]

My rocket scientist professor told me to learn Octave during my master's degree ("it's free MATLAB").

So I watched 197 YouTube videos to learn Python instead.

94.4% of them were a complete waste of time.

But these 11 taught me more than all my PhD professors combined:

[Grant Sanderson @3blue1brown]

New video: But what is a convolution?

https://t.co/gMLxt6uzaf

[Wendy Feng @wendyfyx]

Presented this virtually at @SIPAIM_Society conference last week, and I’m deeply honored that our paper received the 🏆Best Virtual Paper award! Thank you to all collaborators and conference organizers 🎉
Celebrating this weekend with a Porto’s run just in time for Thanksgiving