Been thinking about Ai more than usual this week, and the headlines I see are mostly either AI is going to take everyone's job and is basically an extinction event for most workers, or AI super abundance is coming, and magically ubi will follow.
Ultimately, who knows, there will be a shift, and we'll strap in for the pounding.
But where is the AI cancer hype (or maybe my algo isn't feeding it to me yet), like it seems like a no-brainer that ai models can be used at least in all kinds of diagnostics and should agment the work of any doc - like show medBOT a picture of your mole and give it other health data and let it cross check against every picture of a cancer mole ever taken.
And what about ai models being deployed in cancer research for detection and coming up with new treatment ideas?
Like every time I see another article about a young person getting bowel cancer, I can't help but think, when are the massive medical breakthroughs coming? I don't know how generally bullish I am on ai, but i am super bullish on medical ai, like let's fucking go.
Bonus sats to anyone sharing some good examples of some wizzard level shit ai is already doing in this field.
It might "cure cancer" by telling people to stop doing all the shit that leads to most cancers.
i think most people know the basics but i also think that most people just wouldn't follow the advice anyway
I know. In my view, cancer, heart disease, and diabetes have essentially been cured already because we know how to prevent almost all cases of those diseases.
Just because people won't follow the treatment, doesn't mean the treatment doesn't exist.
How does one prevent all cancers?
I didn't say "all".
Eating real food and avoiding environmental exposure to known carcinogens will prevent more than 90% of cancers. There are still a small number of cancers that seem to be based on underlying genetics and we don't know how to prevent those, yet.
I don't think that's right in my experience. A small percent is genetic (we usually see these early). Then after 50 years old or so it's mostly bad luck. I've seen plenty of people who do everything right who end up with bad cancers- never smoke, marathon runners, eat clean etc. Which is why motto is anything in moderation- you can optimize for whatever fad you think think prevents X Y or Z but end up with W or Q disease or hit by a bus. Not worth the stress.
My guess is 75% is a combination of genetics, bad luck and time. Your DNA undergoes countless replications and repairs after decades and shit happens after a while.
How much have age-adjusted cancer rates changed over time?
We know our underlying genetics aren't significantly different, so either that change is due to our luck changing or it's environmental.
I don't know. The more cancers they cure, the more UBI recipients.
And the more Bitcoin goes up
#1441983
#1439763
labor omnia vincit
If I remember correct it was around here when Dario said something like it would be hard to predict for tasks that are unverifiable. I'd put cures for cancer in that category.
Dwarkesh Podcast: Dario Amodei — "We are near the end of the exponential"
Starting from: 00:12:40
Episode webpage: https://www.dwarkesh.com/p/dario-amodei-2
Media file: https://api.substack.com/feed/podcast/187852154/cecdbe38125e2786cbfebe31dd083d4f.mp3#t=1016
AI might identify potential drug candidates sooner but you still have to do the randomized controlled trials to see if there's a real benefit which is costly and takes TIME to enroll people in studies.
There are so many drugs that have been hyped up in molecular or mice models that end up showing no benefit or even harm.
I thought the same thing about CRISPR. Remember allll the CRISPR hype like 15 years ago? It was gonna be so revolutionary etc etc. As far as I know they've used it for like one thing.
cancer is very easy to cure, just eat zero or less sugar. Learn about your methabolism and DYOR.