aspuru.bsky.social
Professor of Chemistry and Computer Science
University of Toronto
Faculty member, Vector Institute
Director, Acceleration Consortium
Senior Director of Quantum Chemistry NVIDIA
My views expressed here are personal and are not those of my employers.
274 posts
5,580 followers
1,171 following
Regular Contributor
Active Commenter
comment in response to
post
Dude, I think just benchmark on absolute performance. Don’t go that rabbit hole.
comment in response to
post
In TARTARUS we think of it as number of oracle calls: arxiv.org/abs/2209.12487
comment in response to
post
Hard to measure “compute time” eg do you have GPUs or not? How to compare across setups, etc.
comment in response to
post
Thanks for publishing with us! @digital-discovery.rsc.org
comment in response to
post
Russia ? 🤣 China ? 🚀
comment in response to
post
Thank you @jenlove.bsky.social !
comment in response to
post
Hero was epic.
comment in response to
post
Claro que lo hablamos. Y si no nos lo entienden ustedes, solo recuerden que hay más Mexicanos (con X) en el mundo que gente que vive en España. Ojalá te vayas acostumbrando a nuestra pronunciación y nuestras palabras, pues producimos mucha literatura, música y cine. ¡Viva México!
comment in response to
post
Yes. Some of them. And some of them will get jobs abroad. But as the thread is discussing, there is no infinite capacity abroad.
comment in response to
post
There is an assumption here there are enough jobs abroad “waiting for them” as if our other countries are not at capacity. Yes there will be some programs to bring *some* but they will be a tiny minority.
comment in response to
post
2005 - www.science.org/doi/10.1126/...
2010 - www.nature.com/articles/nch...
2014 - www.nature.com/articles/nco...
2024 - arxiv.org/abs/2401.09253
Some of our contributions to the #quantum domain
comment in response to
post
This is true but mass action like a national strike is the only thing that I think could move the needle to stop the dismantling of the U.S. that we are seeing day by day.
comment in response to
post
75-80 percent seems high for Harvard, but regardless, which research dollars are you talking about given the cuts and second with how many strings will they come attached?
comment in response to
post
Not bad choices.
comment in response to
post
I, of course, support the way they are fighting. It is my former institution and I appreciate it a lot. But like CEU, what is the chance of really operating freely under this or any other future neo-Republican regime? If the enterprise is what matters, move it elsewhere. One-time cost.
comment in response to
post
Right move. Ad astra per aspera.
comment in response to
post
Wow I did it on a rush ! Good catch :)
comment in response to
post
Hear from chemist and computer scientist Alán Aspuru-Guzik www.cell.com/news-do/50-i...
@aspuru.bsky.social
comment in response to
post
Glad you will be researching this important question and working with the @accelerationc.bsky.social @kmcelheran.bsky.social !
comment in response to
post
Who? Without their names it doesn't count. If you don't want to do it in public, I expect a confidential DM.
comment in response to
post
Another mistake would be to declare the conclusions false. There is growing research showing how AI and automation may enhance scientific discovery - it's just not in the econ journals. I recommend a Google Scholar search on "Self-Driving Labs"...
comment in response to
post
I took advantage of one of them 8 years ago, the Canada 150 Research Chairs program which was designed for this purpose.
comment in response to
post
More coffee for the students, just electric current and GPT and Claude Credits to get El Agente happy.
comment in response to
post
I agree with Jan! Never imagined that LinkedIn will be useful, but upon the demise of Twitter, it became the home for sharing science. I like Bluesky but it is more quiet now. No tricks from me, Mario.