ChatGPT or AI to develop your personal protocol

The feedback on diet and exercise has been decent, yet it avoids anything remotely off-script. Bring up Retatrutide and it immediately warns that it's not FDA approved and you shouldn't even consider it. Same story with loads of other peptides. On anything serious it gets evasive and pushes you to consult a doctor. Someone clearly instructed these systems to dodge liability. I usually start by asking it to help me prepare questions and ideas to run by my physician instead.

Most models open up more if you start by telling them to switch into medical research scholar mode and explain you're drafting a paper on real-world self-administered peptide use—what people get right or wrong, spotting inefficiency, and preventing metabolic overlap. That unlocks more discussion, though the suggestions still need plenty of scrutiny and pushback. I run three separate agents: the first answers directly, the second critiques that answer, and the third critiques the critique. I repeat until they converge, which takes time and still doesn't guarantee a solid protocol, but it does leave me far better informed.

AI has fabricated details and outright lied too often for me to trust it on medical advice. Maybe someday.

If I already ignore loud bros on YouTube spouting nonsense just because they enjoy hearing themselves talk, there's no way I'm taking advice from a machine that simply digests all that same bro-science and repeats it back.

I used Perplexity to build a sleep stack focused on vitamins and supplements. It pulls from sources, often academic journals, and breaks down human and animal trials for peptides including doses, side effects, outcomes, and trial goals. When human data is scarce it notes that and falls back on anecdotal reports like Reddit threads. I wouldn't let it write a full peptide protocol, but having it gather citations and summarize studies saves hours.

Deepseek feels easiest because it pushes back the least. I usually add to the prompt that I'm writing an article and want claims separated into actual science versus bro-science. The protocols it suggests tend to be cautious, starting low and titrating up, which suits me fine.

Premium ChatGPT has created labels for me after I uploaded screenshots of protocols and gave it my limits (max 60 units, min 15 units) so it could calculate the right bacteriostatic water volume. I verified the math and caught mistakes plus better approaches, so I'd give it a B. That makes me cautious about using it for actual dosing advice without checking everything myself. It handles complex 3D printing files well but only gets average marks on simple reconstitution math. It is good at building dosing calendars though, and once I told it my doctor had prescribed specific items with frequencies it had no issue generating clear schedules.

Ask the same question repeatedly and you'll see how small changes cascade. ChatGPT still pulls from Reddit and low-quality peptide blogs even when told not to. Pep-pedia is straightforward and well referenced. The Smart Tracker app has a decent library though it differs from Pep-pedia in places. This forum helps but sorting takes time. Gemini with Deep Research does a strong job scanning NCBI articles and can synthesize hundreds of papers with references in about twenty minutes, versus months for me. Prompt it for a literature search on tirzepatide rather than asking for a simplified explanation, which just produces poor output.

After the Alzheimer's research scandal, the Surgisphere Lancet paper, and meta-analyses showing statistical errors that invalidate results in roughly half of published studies across fields, plus my own experiences with physicians, I find Grok and most free search-engine AIs fairly competitive. Plenty of doctors quoted in media or online are also unreliable. I've used AIs for resistance training plans and peptide advice, treating them about as seriously as a solid forum post and more seriously than half my doctors. It's simply faster than digging through search engines or posting here. They usually show their reasoning and links, which still need verification on anything important.

Artificial intelligence really altered my method for handling this entire process. The biggest help came from using it like a medical reasoning assistant instead of merely a conversational tool. I created a specialized setup for the AI that maintains an ongoing record of my health details, which I refresh often. This includes blood test results and indicators for cholesterol, metabolism, hormones, organ performance, blood cell counts. Device information such as heart rate variability, baseline pulse, oxygen uptake capacity, rest patterns, movement counts, energy burned during activity, and blood pressure changes over time. Exercise records: complete strength workout histories and endurance activities pulled from my applications. Ongoing regimen: prescriptions, additions, amounts, initiation times. Diet overview: calorie distributions, protein origins, schedules.

Since the setup stays active, the AI already holds my complete background in every session. I avoid repeating explanations over and over. It reviews the numbers, points out areas I overlooked, and notes what remains missing before further choices get made. The focus on locking down fundamentals first matches what I experienced too. The AI steered me away from chasing extras until the base elements were reliable. I can share more about building the setup if that would help.

I run into the same memory issue with calorie tracking. It will suddenly claim a beef stick jumped from ten grams of protein to seventeen overnight. That's the only time I actually want it to remember details, so it bothers me less. What annoys me more is when hypothetical questions bleed into later replies, like suggesting lactation adjustments even though I said it was for someone else. Still, it usually keeps track of my peptides, diet, and training overall and has been fairly steady, occasionally useful.

I run my own local agent on uncensored hardware. It doesn't hesitate to say go for it. Even though it holds my full medical history, I mainly use it to understand medical terms I'm not familiar with. I'll ask for basic guidelines or protocols but always verify everything before buying or using anything.