MD PhD student here and this is 100% on the money and what many I work with try to do. The diagnostic classifiers seem to be about pushing papers and getting funding.
Are places generally adding more than the recommended level of fluoride in their water? How much water do you have to ingest to reach the maximum recommended exposure? Given how bad people are at drinking the recommended level of water, I'm curious how they accurately assessed outcomes based on what I assume are relatively few data points
I've been working on a PR to document and make atomic pointer interface public in Julia. It's really not that hard to handle it under the hood so that array indexing and translates to LLVM and C zero indexing.
Its the one genre of music that I just don't like. I don't like smooth jazz but I've heard songs I can get into. But country music is different. I even spent a couple years putting a dedicated effort into trying to be tolerant of country music. It didn't grow on me.
I erased a lengthy response I was writing because I think our sentiments may actually align. In short, I'm about as certain that these computers aren't going to achieve intelligence in my lifetime as I'm certain that the world is round. But none of that means these models aren't dangerous. Misinformation and accelerating backdoor access to network infrastructures are a couple that we are already see happening.
The lack of efficiency in the brain is a trade off for adaptability. There's no doubt that we can transmit signals faster than neurons can fire but the cost is drastically reduced adaptation. There are micro, meso, and macroscopic networks in the brain that have different degrees of "adaptability". This isn't even considering the variety in neurons or the additional signalling cascades by non neuronal tissue. How does all of this contribute to intelligence? We don't know exactly but much of this has survived millions of years of evolution in many animals so it probably has some role.
That's not to say we can't do this with computers and less computational power. However, it's really improbable that a couple layers of adaptability on an artificial neuron network will be anywhere near sufficient to simulate intelligence in even a rodent.
Love the visualizations. It would be great to have `realloc` covered too but I'm not sure how much that varies by system, potentially making it too complicated for this sort of post.
That last bit is how I sometimes feel about figuring out how the brain works compared to casually learning computer science stuff on the side. I love being a part of improving our knowledge of how the brain develops and sometimes goes wrong, but it's way easier to debug obscure source code than the brain.
Running MRI processing on a super computer and I want to check at a certain point in the pipeline that things went correctly by viewing a couple of slices without downloading a huge brain image
Amen. There's always that first time where you're able to trick them but then they know and start to realize you sometimes trick them. For us that meant this obnoxious gap between being smart enough to avoid deception and smart enough to appreciate the higher level importance of something like nutrition. I don't know if my oldest ate anything green between the ages of 3 and 6.
I see where you're coming from but we've had a lot of politicians over the last several years implicated in some pretty bad stuff. From sexual harassment to sex trafficking.
Refinement types are pretty interesting and I've seen mention of them recently. I'm curious if there are any practical reasons we don't see them implemented in more languages.
I didn't even know some of these things were being worked on until recently. I totally understand why devs don't treat development like a Twitter feed, posting every thought that pops into their head instead of working. However, it would be really interesting to follow some of these developments without having to deep lurk all the PRs.
From a non-technical point of view (since the technical answer was already provided) I think this sort of magic optimizations is a double edge sword in any language. No matter what you do there will be corner cases that need manual tuning that become inaccessible behind some init option.
If you frequently want to develop and maintain publicly uses functionality that requires writing some in a faster compiled language and then binding to an interactive one like R or Python. Test coverage and multiuser maintenance is way easier when it's all just one language and has a sud package manager.
Saw elsewhere you were on PC. I'm on a series x and it plays really well. I don't know how much time I've spent on it but it was way too much because I was getting back into a C++ project and played it whenever I had long compilations. I only had a couple of crashes and it was about 6 months ago too, so it might just depend on what systems they are fixing bugs for at a given time.
That being said, once you finish everything it really does feel empty. There's no repeatable content once the game is finished, unlike most other open world games. This is pretty unfortunate because I would sometimes just happen upon fun little notes and hiding spots. I'm sure there are tons of little details that will never be noticed because there's no drive to go out of your way and find them once all the other stuff is done.
I still remember sitting on my father's lap as a small child while he played Wolfenstein on the computer. Then my mom got mad at him for letting me watch him play.
I don't know the specifics in this situation but I work at a university where vaccines are actively developed (for many things). Unless there's a significant financial push, administration to children lags far behind to ensure that random autoimmune complications with vaccine are far below the risks of catching the virus and having a serious reaction.
Also, you can pretty much anticipate pregnant women will be the last two be approved for anything. Pregnancy complications are common enough that no one wants to mess with that whole situation until you're certain you won't have cross antigen reactivity in the host
You're not wrong but who is going to read a software devs blog? If this was some major news outlet writing a tech piece then the public at large may actually be reached with this info
We don't really do brain mapping in the sense that would apply to nanotechnology. The actual mechanism of working memory is pretty hard to establish in humans at this level.
We use this in my lab and I think you this is a lot more complex than better or worse on the task as a whole. Certain kind sof stimuli will interact with subject memory in different ways. So even if there's research saying another species is better or worse it probably depends on what is being recalled.
This is interesting for basic research but this strategy is unlikely to ever make it to human research. We need to find ways of extracting more information at reasonable Tesla. 9T just isn't gonna make its way to clinic
So true. There are still plenty of areas where we lack sufficient data to even approach applying this sort of model. How are we going to make similar advances in something like medical informatics where we not only have less data readily available but its much more difficult to acquire more data
Just by virtue of how the human brain transducers auditory signals versus visual signals there is a big disparity in being able to automate music production. We fill in many gaps for vision but to a much smaller degree for music. There's a lot less room for AI to fudge the difference between human and generative model derived music.
It will be interesting to see how this all plays out as I'm sure AI will found some use in music that has broad application and influence. But it will take a lot longer than generating pictures
This isn't like genome fingerprinting. It's only possible with someone's consent because you have to put them in a machine and accurate fingerprinting like this takes many hours of acquisition before it's reliable. Getting this sort of thing to the same level of reliability of genetic fingerprinting is decades away
Interesting research but some of the writing is a bit misleading. For example, it talks about sleep problems prior to ASD but people aren't "developing" ASD into adolescents.