Emmitt--I appreciated the post detailing your "dispute" with Chat. Revealing... though not particularly surprising.
I was struck by the "caveats" that Chat brought up: leronlimab remains unapproved, Cytodyn is still pre-revenue, there are financing challenges, and "Phase 3 validation in oncology or other new indications has not yet occurred." Change the company name and the drug and you could say that about ANY pre-revenue biotech working on cancer. And Chat did not bring up that leronlimab has successfully completed a Phase 3 trial in HIV. So a lot of caveats but nothing of any significance if you know the company... and downright misleading if you don't.
Something similar when Chat suggests that many believe the company is materially different, yet the "ultimate judgement" will depend on-- reproducible clinical results, regulatory execution, successful trial design, and approval or partnerships. Pretty obvious that ANY biotech needs to successfully address those issues to prosper. Sounds reasonable, says nothing new. I would have asked Chat to evaluate Cytodyn's progress on those metrics to see what they say.
I suspect most of you use AI pretty aggressively, but if not, here's what I do (if it helps):
--Interrogate it. Ask follow ups specific questions. Those of us on this board may actually know more about leronlimab and CCR5 inhibition than AI does, at this point anyway. That won't last for much longer, of course...
--Correct when wrong. For example, my google AI confused Vitamin K1 with K2. Rookie mistake a high school biology student would not make. Seems to learn, so treat it like a student that needs direction.
--Instruct AI to look at specific academic references when getting deep into a topic. So instead of asking AI to "suggest possible biomarkers of metastasis," I might say: "Looking at Pestell AACR 2026, evaluate cancer clusters, CCR5 dots, and CD276, sTyro3, and TNFSF13B as potential biomarkers of metastasis." (That is just an example, by the way. Posing these kind of questions with reference to recent work by Adams/CreativBio might be very fruitful... That's where I think Cytodyn and Creativ are heading, by the way).
--I actually find some of the "prompts" at the end of a response useful. "Would you like me to look into" this or that. I utilize polyphenols, a carotenoid, and a triterpene (and K2) as over-the-counter therapeutics for atherosclerosis, and found AI prompts helpful in developing a daily protocol, verifying appropriate dosages, and learning about contraindications. For example, pomegranate polyphenols bind to non-heme iron in the gut, so if you drink P juice with steak or spinach the iron will not be absorbed. So I avoid iron-rich foods when drinking my juice, and watch my iron levels. And the punicalagins and punicic acid in the P juice will also go to waste, instead of stripping away the plaque in my arteries. So I learned something new by letting AI take me there. I don't necessarily "have it out" with AI, like you did, Emmitt... but I try to work with it if it is leading me somewhere useful or interesting... Yet I do take a kind of perverse pleasure in correcting AI when it's wrong... And it apologizes!
Glad you had it out with chat, and posted the results.
I was struck by the "caveats" that Chat brought up: leronlimab remains unapproved, Cytodyn is still pre-revenue, there are financing challenges, and "Phase 3 validation in oncology or other new indications has not yet occurred." Change the company name and the drug and you could say that about ANY pre-revenue biotech working on cancer. And Chat did not bring up that leronlimab has successfully completed a Phase 3 trial in HIV. So a lot of caveats but nothing of any significance if you know the company... and downright misleading if you don't.
Something similar when Chat suggests that many believe the company is materially different, yet the "ultimate judgement" will depend on-- reproducible clinical results, regulatory execution, successful trial design, and approval or partnerships. Pretty obvious that ANY biotech needs to successfully address those issues to prosper. Sounds reasonable, says nothing new. I would have asked Chat to evaluate Cytodyn's progress on those metrics to see what they say.
I suspect most of you use AI pretty aggressively, but if not, here's what I do (if it helps):
--Interrogate it. Ask follow ups specific questions. Those of us on this board may actually know more about leronlimab and CCR5 inhibition than AI does, at this point anyway. That won't last for much longer, of course...
--Correct when wrong. For example, my google AI confused Vitamin K1 with K2. Rookie mistake a high school biology student would not make. Seems to learn, so treat it like a student that needs direction.
--Instruct AI to look at specific academic references when getting deep into a topic. So instead of asking AI to "suggest possible biomarkers of metastasis," I might say: "Looking at Pestell AACR 2026, evaluate cancer clusters, CCR5 dots, and CD276, sTyro3, and TNFSF13B as potential biomarkers of metastasis." (That is just an example, by the way. Posing these kind of questions with reference to recent work by Adams/CreativBio might be very fruitful... That's where I think Cytodyn and Creativ are heading, by the way).
--I actually find some of the "prompts" at the end of a response useful. "Would you like me to look into" this or that. I utilize polyphenols, a carotenoid, and a triterpene (and K2) as over-the-counter therapeutics for atherosclerosis, and found AI prompts helpful in developing a daily protocol, verifying appropriate dosages, and learning about contraindications. For example, pomegranate polyphenols bind to non-heme iron in the gut, so if you drink P juice with steak or spinach the iron will not be absorbed. So I avoid iron-rich foods when drinking my juice, and watch my iron levels. And the punicalagins and punicic acid in the P juice will also go to waste, instead of stripping away the plaque in my arteries. So I learned something new by letting AI take me there. I don't necessarily "have it out" with AI, like you did, Emmitt... but I try to work with it if it is leading me somewhere useful or interesting... Yet I do take a kind of perverse pleasure in correcting AI when it's wrong... And it apologizes!
Glad you had it out with chat, and posted the results.
5
