Timothy Oh

neuromorphic computing, machine learning, theoretical neuroscience, complexity theory, physics

[Timothy Oh @alskaf1293]

my head has too many ideas running through it 24/7

[Timothy Oh @alskaf1293]

@ABhargava2000 wrote a blog on my work. check it out: https://t.co/mbMqiaEDow

[Naveen Rao @NaveenGRao]

Computer science is a weird field. It’s literally an entire field based around the operation of a particular machine. It's the equivalent of saying "piston-based internal combustion scientist" instead of mechanical engineer.
There are so many other ways to build a "computer". This field will change drastically over the next 15 years. It's kind of converging back into mathematics as a system of information representation an manipulation.
And yes, I have a degree in CS.

[Timothy Oh @alskaf1293]

@evisdrenova @willfp you could also gather the sand yourself, turn it into pure silicon, and build the entire chip yourself. no need for abstraction!

[Timothy Oh @alskaf1293]

@InnocFrancesco predictive coding is provably more efficient if you instantiate it as a hardware substrate
https://t.co/kYK4zY0DxE

[Timothy Oh @alskaf1293]

@ptolemyofsinope @ABhargava2000 talked to him a year ago and we quabbled a bit over definitions. i think it’s good to know as well even though the criticism that it’s unfalsifiable is fair

[Timothy Oh @alskaf1293]

@neural_avb interesting result theoretically, limited uses practically

[Eric Weinstein @ericweinstein]

If you are not willing to kill or die for your own most treasured values, you should think carefully before welcoming in migrants who would do both for theirs.

[Timothy Oh @alskaf1293]

the prospect of space exploration is so interesting
i think we can have some sort of thing where we first get launch costs down (spacex's rocket garden) and develop an economy in space
solar farms would first be sent up to provide energy for machines to do work and move stuff around -> asteroid mining -> in space refining/manufacturing -> more solar based solar power -> cycle continues
could launch us to type 1 by the early 22nd century

[Timothy Oh @alskaf1293]

I took a class under a Nobel laureate. For one class we were placed on the outskirts of campus, and he emailed the school and it was solved in a couple hours - no bureaucratic friction. Every conference that he goes to there’s always some people that want to take pictures or talk to him in some capacity. He also has a parking spot reserved for him close to the university. Might interest you 😂

[Timothy Oh @alskaf1293]

@pogarithm more ideas are to come with computers. contrarianism still exists but in the lone individuals

[Timothy Oh @alskaf1293]

@JosephJacks_ both effective theories, at two completely different scales

[Timothy Oh @alskaf1293]

@fossilflavoured Folgen von ADHD 😔

[Timothy Oh @alskaf1293]

There was this language created in the States called Ithkuil which was designed to be a "maximally dense" information language. The idea was that speakers could communicate entire sentences worth of meaning in English in just a phrase in Ithkuil. We should just speak that to the LLMs

[Timothy Oh @alskaf1293]

@ProfStanciu Nietzsche's last man

[Timothy Oh @alskaf1293]

The history of invention has been one of value creation, not value capture.

Tesla is the canonical example. He's credited with inventing AC power systems, motors, and radio systems while Westinghouse captured the value.

Tim Berners Lee is credited for the invention of the World Wide Web, while the companies that formed due to this invention (Apple, Facebook, Google) captured the value.

[Timothy Oh @alskaf1293]

@cyb3r_17 the future is neuromorphic. the instantiation of a new computational class

[Timothy Oh @alskaf1293]

@SchopenhauerNow Nietzsche

[Timothy Oh @alskaf1293]

@gabriberton look into predictive coding, contrastive learning, and equilibrium propagation. All approximate backprop while having more biologically plausible features

https://t.co/odXEhcvOKc