Forum Replies Created
-
Posts
-
<span class=”bbp-author-avatar”>
</span><span class=”bbp-author-name”>Fiesty</span>
Thank you for taking time to write this out, I’m finding it just takes time and patients, but worthy the exploration. You made it easier and worthy the effort.
Thank you for the referral to the substack article. I rather like his first bullet point: “LLMs are a flawed technology”, great place to begin. This makes sense because the same can be said of us humans as well, always a work in progress to be certain.
We say we’re committed to information and truth while fighting it every step of the way. A good example from my work as an RN is that much of what was imposed upon society made little sense, or at least that was the “inside baseball” discussions as we fought to do what we can within the veil of ignorance of the first months of the pandemic. People said they want to “know”, but didn’t want to hear.
The previous views regarding the value-added process of information, missed the first and last steps, the latter of which we face today. I would re-write this to look like: Noise -> Data -> Information -> Informing knowledge -> Insight -> Productive Knowledge -> Action. Having ubiquitous access to knowledge still requires, at least today, some application of human insight, often experientially acquired. AI may provide the correct answer, but so far not the proper action.
Insight feels more domain specific as well. I may be able to provide insight into a patients future post heart attack, but I would fail miserably in structural integrity of buildings or bridges. I don’t see how experiential insight is accessible through LLM’s given they are trained on what we say among ourselves.
In the end, it feels like the superpower required to navigate a world supported by AI is a mixture of humility and good self-esteem. The former keeps us open, the latter is necessary when decisions need to be made where certainty of outcome is not possible. Fun times, don’t you think?
Thanks again for the reference, perhaps we scooted down the road a bit toward understanding our role in all this.
//tom
App Store will be launched over the next few weeks. This one will be developmentally complicated because it will involve revenue sharing and all that needs to be built. AI will make that faster, still lots of testing required. I’m building one for a friend of mine recently diagnosed with pancreatic cancer the point of which is to discover questions to bring to her oncology team. She too is an RN so we’re working it together, she just doesn’t know it yet.
Johan, Thanks for the assist! I added that to the prompt this morning, great idea. Here are the choices of access inside the configurator, public isn’t enabled yet I changed this to Public (image below) to see if that changes behavior, https://chat.openai.com/g/g-IRYUvWHbK-longevitylab
I worked through my undergraduate in Economics as a critical care and emergency room RN, so my biases are focused on lowering barriers to entry in acquiring simple clinical information. As you can imagine, when people face clinical challenges, general information about their condition is helpful, but far from sufficient. We need that general information to be informed by who we are by age, gender, height, weight, BP and many more. Here is where the bulk of medical expenses are incurred. I’ve seen friends of mine at Church dealing with complex medical conditions and it’s like: “well here’s your diagnosis, now go to graduate school and learn about it”. My LongevityGPT is just that without all the library time, or at least that’s the idea.
Thanks again for the recommendation. This is an emergent technology, we couldn’t ask for a better time to be around. I’m 70 (geese that’s the first time I’ve typed that, kinda weird) with a biological age of 65 so I hope this Longevity stuff keeps me around so I can enjoy the end of the movie.
I’ve made a few, quite amazing to be certain. Brian’s work has helped greatly, as has Multiplex.
I work with patients over 65 and help them in reducing their biological age and trying to decentralize it to a GPT. Here’s an example with much more to build: https://chat.openai.com/g/g-IRYUvWHbK-longevitylab. Might be a good time to short medicine. 🙂
Johan,
Thanks for bringing this up, I’m with you. I’ve been paying, which has been helpful in understanding how plugins work, still there is so much I don’t know, and more I don’t know I don’t know. It would be great to fire up a Zoom huddle with folks working through the same conundrum.
Thanks, //tom
Kenneth, Thanks for adding to this discussion. I’m curious as to how all this will play out. In the world in which I have lived, I’ve always been impressed by individuals with advanced degrees. I enjoy the intellect as well as a society wealthy enough to allow people to do deep study; work that can feel unproductive for years. AI exposes me to that knowledge without the hassle or individual ability to work on complex issues myself.
Will we discover that we need fewer Ph.D.s, or will it make existing Ph.D.s more effective and efficient? I don’t know, still I am quite entertained exploring topics outside my domain for insights into increasing human longevity and thriving. Fascinating times to be certain.
Attached is an image trying to work through what value-added human characteristics will be best advantaged by AI, and it doesn’t look like IQ is as important as it had been. Yea for my team.
Thank you, this helps. If I’m interpreting this correctly, it seems like a “Proximity to Cause” challenge. As we are all familiar, if I were to tap you below your knee-cap and your leg were to move forward (Inferior Patellar Reflex), you wouldn’t be surprised. What if your friends leg moved instead? That might be noteworthy. In complex systems, such awkward surprises happen. At this stage, it appears humans have a distinct advantage in sussing out such unique, non self-evident associations. Are insight and wisdom, post production events? What skills should we point our children toward? What if music proves to be an advantage to wisdom production, or stand-up comedy? What if at the end of all this it boils down to Galatians 5:22-23: “But the fruit of the Spirit is 1. Love, 2. Joy, 3. Peace, 4. Forbearance, 5. Kindness, 6. Goodness, 7. Faithfulness, 8. Gentleness 9. and Self-control; why not?
Thanks again//tom
In the article attached from 1984, was written to explore the noise around “Information Highway”, and what that might mean. This is an amazing piece to which I often refer. We take for granted today what was once new and or speculative. We’ve gone from Noise -> Data -> Information -> Informing Knowledge -> Productive Knowledge -> Action. So far this has worked. It seems now, with AI, we all have access to Informing and Productive Knowledge. We use to depend upon Ph.D.’s and other experts to help us make sense of the tsunami of information we’re exposed every day, and now, it seems, we can cut-to-the-chase and tap into knowledge on our own.
Here is where I need help. It feels like there is another step following knowledge, a new super power that needs recognition and development: Insight. Am I on track here? I need help in flushing this out. Where should I go from here?
Thanks
//tom
These are all excellent, none require a prescription. The most dynamic science in medicine is in the domain of longevity, and it’s amazing. In this AI can be very helpful because much of what individuals can do, they can do on their own in consult with their physician. And we’re talking significant reductions in biologic age, even if starting in the 70’s and beyond:
- Calcium-alpha-ketoglutarate
- Fisetin
- Quercetin
- n-acetyl cysteine
- D3 with K2
- NMN or NR
Rx required:
- Metformin
- Rapamycin
You’re going to see some amazing science over the next 5 years, studies in progress. Just consider the possibility of growing younger, take off 5 to 10 years. Why not?
@qlockmaker, thanks for this prompt. I’ve often turned to GPT4 in helping me with patients, particularly in setting things up so that they can pursue questions as they arise. In some situations, I’ve found that intuition that comes from being an RN, help in designing the prompt. In particular, my 3 year old granddaughter needed heart surgery so I was able to input pertinent clinical findings from multiple reports. What kind of experiences have you had? Thanks again, //tom
Brilliant! Equaling Asimov is a tall-order ask, and the right one. His three rules for robots are great because they are self-referential, they work like the Bitcoin protocol, automatic, immutable. It is troubling when I look for insight into the dataset GBT4 was trained on, it’s opaque. My reading of the math is that biases are self-correcting, though it is difficult to tell.
Trustworthiness strikes me as a central goal that will drive value. So far I’m not seeing anything alarming or intentional, though you could imagine the incentives for mischief are huge which leads me to the human attribute of insightfulness, where knowledge meets time. So far, I see no risk to humans, insightfulness is what we do best, that and miss-read intentions.
//tom
We educate ourselves to become more valuable faster by learning from others. In this way, wealth favors the intelligent, certified by the result of intelligence put to work or institutions of higher education who “certify” that intelligence. With AI, I can access knowledge and expertise by asking a carefully worded question, designing a prompt and probing. In a world of widely available AI, the apex skill is moving beyond IQ to be replaced by Galatians 5:22-23.
22 But the fruit of the Spirit is love, joy, peace, patience, kindness, goodness, faithfulness,23 gentleness, self-control; against such things there is no law.
