In reality, AI will help solve a lot of our medical ignorance. Not the goofball LLMs, but specialized AI algorithms specifically geared towards niche medical research and applications. Don’t let GPT and Gemini sour the potential of some possibly game changing software.
Comment on Breakthrough gel can regenerate tooth enamel within weeks
HugeNerd@lemmy.ca 3 weeks ago
Our medical knowledge is shockingly rudimentary. Why we can’t coax cells to do what they did anyway before is something we really need to understand to pretend to have any kind of medical knowledge. What passes for medicine is nothing more than 19th century++. See this, do that, body will heal. That’s about it. I’m shocked every time I see someone in a wheelchair, how can we have LLMs and hallucinated movies but not understand how a few milligrams of organic matter organizes into nerves? etc etc etc
I am hoping that the absolutely bonkers computing power we are currently wasting on fart videos will be used to simulate matter in the future, here’s to cheering on the AI crash!
MumboJumbo@lemmy.world 3 weeks ago
frongt@lemmy.zip 3 weeks ago
The human body is an immensely complex system of chemical reactions.
And before you say “simulate it”, we don’t have nearly enough computational resources to do that. Simple reactions or chains of reactions, yes, but even a simple body process consists of multiple steps and a number of large, complex molecules.
Doomsider@lemmy.world 3 weeks ago
This right here. We can definitely model how a single chemical effects cells, but what about multiple chemicals. The reality is we are still very far from simulating something like the side effects from multiple interacting drugs at once.
We will likely discover some incredible insights when we do this though.
Jumpropegazing@lemmy.zip 3 weeks ago
we do have the ability to coax cells and even turn differenciated cells into pluripotent stem cells we’ve been able to for years the people who did it won a nobel peace prize
Soggy@lemmy.world 3 weeks ago
We have a pretty good idea how things work during development, it’s tricking those cells into the same process as a fully-formed organism that’s hard. Isolating and distributing the hormones in the right way. I think you’re underselling the complexity and scale of biology. We can’t just put a tiny camera and chemo-sensor inside a neuron and see what’s going on in real time and synthesize up a hundred thousand copies.
HugeNerd@lemmy.ca 3 weeks ago
That’s my point, we’ll simulate it.