Quote:

---

Every satellite, probe or lander sent into space from Earth that is successful demonstrates the "wisdom" that can be gained through simulation. Every failure also gives you wisdom about what not to do next time and that you simulations need to be tuned up.

---

In that scenario the vehicles are used to gain actual data, it's not done through simulation.

Quote:

---

This AI discussion is about evaluating how Leronlimab might work against various diseases, say, on your list. An AI could be used to aggregate all the relevant knowledge we have from research already done. Then it can suggest diseases to target for further investigation owing to the constellation of cells and cytokines and receptors involved.

---

It can suggest, which is quite different from what you stated. Any mistakes could have catastrophic consequences.

Quote:

---

You do realize that AIs will eliminate the need for all drug trials - mouse, monkey and human? All drug trialing will be done virtually.

---

Quote:

---

And that is exactly what is done in experiments in petri dishes - simulations of what might happen in a mouse or a monkey or a human being. You don't think a simulation at that level can be done by an AI?

---

What might happen is not necessarily what will happen.

Quote:

---

But you, Ohm, seem to think that an AI drug trial through simulation would not go through a series of step, going from the most simple to the most complex, but would go from simulation to market. That would indeed be a disaster. No one is suggesting that. That is fiction.

---

Even if AI followed a pre-clinical, phase 1,2,3 model it would still be AI and thus not be real world provable.

Quote:

---

Patterson has already used AI to find the cytokine patterns associated with Long Haulers and CCR5 blockade.

---

In Patterson's case it seems to be buzzword hype, which happen quite often these days. In other words his "AI" uses commonalities between patients. Something that can be done by just looking at a list of biomarker results.

Quote:

---

Dr. Patterson created a blood test that uses biomarkers to diagnose patients. "And we looked at over 150 different biomarkers, and found with the help of AI, a signature that says, this is long COVID."

---

There is a role for AI. It can look at a very wide variety of data and come up with possible solutions. They're only possibilities and quality of data can skew things quite badly. If you've read stacks and stacks of medical papers you'll realize that often they come to completely opposite results so which data does AI rely on? AI is not actually intelligent. Computers are binary and any incorrect data will throw results off completely. The more complex the problem the greater the chance for error. When it comes to human lives putting trust in AI is an abnegation of responsibility.