Announcement The Forum is transitioning to a subscription-based membership model

[Regina]

When people are locked into their 15 minute smart cities, with their smart appliances and GMO foods, they will finally realise that their inability to oxidize glucose is political.

Great quote.

 

[Peatress]

@Peatress

"Agenda 2030 is no small thing."....Especially when one considers they have Agendas 2040, 50, 60, etc, all the way out to 2100...no joke. Overcoming the effects of the covid poisoning will be a monumental task so maybe the forum can be a key resource in the effort to heal as many as possible?

There is a thread on shedding remedies. Maybe we need a thread for people who have successfully reversed vaccine injuries.

 

[Lejeboca]

Yes, maybe that can be made possible with a curated chatbot in this forum giving the better answers in order to point them in the right direction.

This would certainly be more than worth the subscription fee as outside the forum it's a veritable jungle of misdirection and misinformation. with Google and all.

But the chatbot should limit itself to health matters, specifically metabolism and Peat-centered, and shy away from political matters, lest it incur the ire of the censors.

Then charlie would have another problem, albeit a better problem, when a long queue forms in the application for membership line.

This would help set us up to use good information to begin with, and it will be easier for each member to connect the dots in unraveling the mystery that is their diseased condition, and to get better.

The chatbot could compile a list of resources buries within the website including, but not limited to RP newsletters, books, articles, interviews, and the works of scientist often cited such as Szent-Gyorgi, Warburg, Ling, Koch, Metchnikoff.

A member would have enough material to embark on a deep dive for his own particular context, and emerge more knowledgeable, and become more useful as a poster that adds value to threads. The quality of the forum improves with more knowledgeable members.

---____________---

Could someone call Elon for a turnkey Chatbot server system? Tell him he gets a free lifetime membership to go along with it. And free advice from us all.

I like your line of thought. Way to go: a peaty chatbot to jumpstart a deeper research.

Maybe having all Peat's newsletters as a dataset for such a chatbot is a good start?

 

[yerrag]

I like your line of thought. Way to go: a peaty chatbot to jumpstart a deeper research.

Maybe having all Peat's newsletters as a dataset for such a chatbot is a good start?

We're getting good vibes.

Yes, the newsletters alone has enough data to begin with. The chatbot may also give us a good account of how Ray's positions have changed over time, as part of his process of discovery.

The chatbot may even be used as a fact check before a new post is made. It can be a button that the poster can use to review his post before he actually posts it. But using this feature is entirely at the discretion of the poster, as he may be intending to post anti-Peat ideas that he feels he has strong arguments for. And this would keep the forum from sliding into becoming an echo chamber. It would however help in keeping posters from misrepresenting Peat and help the spread of ideas that with each day forms a good nucleus of coherent ideas to connect dots with towards a eureka moment for members in search of answers to their unique context.

With this feature, members would be thrilled to stay as their experience becomes more dopamine-filled. New members on an their first month (a trial membership, if you will) will be convinced to convert to a yearly membership. Grandfathered members on free membership will gladly upgrade, and their support will become in turn not only be about providing useful content but also in keeping the website financially robust.

Over time, the content of this website increasingly becomes golden and as membership grows, it becomes a magnet for advertisers. This further helps the site provide more relevant content, search, and AI capabilities, as positive testimonials from members builds on how this website has made a huge difference in improving their health.

As its success may attract the scum from its periphery intent on shutting it down, it would have by then developed enough financial strength and goodwill (especially from influential people) that any attempts at extermination by interlopers would be met by a firestorm in social media.

 

[yerrag]

I know, I can be a bore when I get ahead of myself.

Slap me. I must be dreaming.

 

[Lejeboca]

I know, I can be a bore when I get ahead of myself.

Slap me. I must be dreaming.

If these are dreams, they sure feel like vivid dreams!

I am interested to get @charlie 's general feedback, as the forum owner. Charlie, do you think such a chatbot is feasible for the forum and is a good thing to have?

 

[imcoconut]

Yes, maybe that can be made possible with a curated chatbot in this forum giving the better answers in order to point them in the right direction.

This would certainly be more than worth the subscription fee as outside the forum it's a veritable jungle of misdirection and misinformation. with Google and all.

But the chatbot should limit itself to health matters, specifically metabolism and Peat-centered, and shy away from political matters, lest it incur the ire of the censors.

Then charlie would have another problem, albeit a better problem, when a long queue forms in the application for membership line.

This would help set us up to use good information to begin with, and it will be easier for each member to connect the dots in unraveling the mystery that is their diseased condition, and to get better.

The chatbot could compile a list of resources buries within the website including, but not limited to RP newsletters, books, articles, interviews, and the works of scientist often cited such as Szent-Gyorgi, Warburg, Ling, Koch, Metchnikoff.

A member would have enough material to embark on a deep dive for his own particular context, and emerge more knowledgeable, and become more useful as a poster that adds value to threads. The quality of the forum improves with more knowledgeable members.

---____________---

Could someone call Elon for a turnkey Chatbot server system? Tell him he gets a free lifetime membership to go along with it. And free advice from us all.

I'm actually looking into doing this. I'm looking for the most cost effective option, but it's not technically too hard with a lot of ray's writing being available digitally, OpenAI's API and/or the current supply of solid open source language models. The open source tooling in this space is advancing so quickly. It just might be expensive to train and host depending upon much use it gets.

I was also thinking about a separate forum based one that indexes all the posts, but this would be much more expensive given the higher quantity of text.

If anyone would be interested in chipping in let me know. I have a lot on my plate at the moment but I'm going to estimate how much it would cost using different options, with OpenAI being the best (most likely).

 

[akgrrrl]

I know, I can be a bore when I get ahead of myself.

Slap me. I must be dreaming.

Are you kidding? THIS is how the old forum operated. Members like you taking a topic and doing a deep dive on aspects, factors, and possibilities. Then, thinking out loud, unafraid to hear dissent or discussion, or debate. The crisis of the world today is that 95% of humans cannot rub a few words together for a coherent sentence. You are giving us all ideas we can collectively approach, contribute to, and perhaps give Charlie a well-thought out avenue of application to consider. I remain grateful for your willingness to present these ideas.

 

[LA]

Are you kidding? THIS is how the old forum operated. Members like you taking a topic and doing a deep dive on aspects, factors, and possibilities. Then, thinking out loud, unafraid to hear dissent or discussion, or debate. The crisis of the world today is that 95% of humans cannot rub a few words together for a coherent sentence. You are giving us all ideas we can collectively approach, contribute to, and perhaps give Charlie a well-thought out avenue of application to consider. I remain grateful for your willingness to present these ideas.

YOU GO GIRL - not just the forum but the USA in general was once free to speak.
Charlie has a lot of work to do, He is always thinking.

 

[akgrrrl]

YOU GO GIRL - not just the forum but the USA in general was once free to speak.
Charlie has a lot of work to do, He is always thinking.

Grazie. (Curtseys)

 

[akgrrrl]

I'm actually looking into doing this. I'm looking for the most cost effective option, but it's not technically too hard with a lot of ray's writing being available digitally, OpenAI's API and/or the current supply of solid open source language models. The open source tooling in this space is advancing so quickly. It just might be expensive to train and host depending upon much use it gets.

I was also thinking about a separate forum based one that indexes all the posts, but this would be much more expensive given the higher quantity of text.

If anyone would be interested in chipping in let me know. I have a lot on my plate at the moment but I'm going to estimate how much it would cost using different options, with OpenAI being the best (most likely).

Just a suggestion because I have looked up things here---Sayer Ji accumulated decades of data (in many forms) and arrived at GreenMedInfo. I wonder if that could be reviewed for modeling purposes?

 

[Orome]

Subscribed .
I cannot post that much and want to contribute my share to support the forum and keep the legacy of Dr. Peat alive.

 

[yerrag]

If these are dreams, they sure feel like vivid dreams!

I am interested to get @charlie 's general feedback, as the forum owner. Charlie, do you think such a chatbot is feasible for the forum and is a good thing to have?

We're building a field of dreams!

I'm actually looking into doing this. I'm looking for the most cost effective option, but it's not technically too hard with a lot of ray's writing being available digitally, OpenAI's API and/or the current supply of solid open source language models. The open source tooling in this space is advancing so quickly. It just might be expensive to train and host depending upon much use it gets.

Glad to hear you've done some groundwork, at least in your mind. How mature is OpenAI? Is it at the level of v1 already? I've really not kept up much with this field and I'm asking only out of ignorance.

Is there already a prototype, even not fully functional, that is is already at work that you're involved in? Would be interesting to see, if it's not under wraps.

I was also thinking about a separate forum based one that indexes all the posts, but this would be much more expensive given the higher quantity of text.

A separate forum seems unnecessary, and it would be dilutive member-wise. A chatbot within this site would enhance and attract membership.

What kind of server at the minimum is needed to begin with , and would it be easy to scale up the capacity of the server? I imagine a lot of processing capability is needed, but not a whole lot is the chatbot is limited only to members, and even less if limited to paying members.

If anyone would be interested in chipping in let me know. I have a lot on my plate at the moment but I'm going to estimate how much it would cost using different options, with OpenAI being the best (most likely).

Let us know. It would be very helpful to know. I'm just thinking out loud. Probably build best of class reliability is tops hardware-wise. Opensource with a good team behind it, with good track record in other earlier projects, barebones functionality that suit minimum reqt's for satisfactory performance. It is assumed that the ability to input text information would come easy from the existing repositories (such as of @Eth 's). Text-based data would be enough to begin with.

If the server design is scaleable, you can give an estimate of a starter server unit with the minimum needed processors and memory for a limited usage level, and probably open that server up later for limited scale operation with a few users. If we can say get donations from 50 users, how much would each contribute as a one-time donation? From 100? Just trying to keep it realistic as I have no idea. It sounds promising but a lot depends on meeting the expectations we have come to expect from ChatGPT and is that the level we are shooting for? I think it may have to measure up to that, and the cost may be higher to get to that level, but it may be that anything less would be a no-go. But maybe not. I think it is a matter of settig expectations and whether the level of service still meets a minimum that gives satisfaction to each member that uses the service.

Worth going through the exercise though.

 

[yerrag]

Are you kidding? THIS is how the old forum operated. Members like you taking a topic and doing a deep dive on aspects, factors, and possibilities. Then, thinking out loud, unafraid to hear dissent or discussion, or debate. The crisis of the world today is that 95% of humans cannot rub a few words together for a coherent sentence. You are giving us all ideas we can collectively approach, contribute to, and perhaps give Charlie a well-thought out avenue of application to consider. I remain grateful for your willingness to present these ideas.

Thanks again for the support. Anything we can do to support Charlie and in that way ourselves as well.

I've never invited any friend or family to this site, for many reasons. Chief is they are used to the convenience of not having to dig for answers. And in our forum, we are used to digging. So rather than invite them to dig, I rather not. But they have less digging to do with AI well implemented here from the get go. I would send them a month membership (that's like a trial membership- not available just so you know) and they'll be floored by the experience, and become a paying member at the end of the trial.

Ray Peat has the best ideas, and the best answers. I truly believe that. But the 99% believes in the non-Ray Peat pagan health religions. Converting them is hard, even with AI, but if you can convert the 1% from the 99% with it, that is a good start.

 

[charlie]

I am interested to get @charlie 's general feedback, as the forum owner. Charlie, do you think such a chatbot is feasible for the forum and is a good thing to have?

I am already working with a programmer but do not want to get into the specifics right now.

 

[ThinPicking]

The open source tooling in this space is advancing so quickly. It just might be expensive to train and host depending upon much use it gets.

Privately training and hosting something based on an "open" model like LLaMA is going to be cost prohibitive for "free" (ad funded or otherwise) public use where a reasonable response time's expected. A subscription model could work but continually tracking its use, making projections and rescaling it would be tricky to say the least. For private use it's absolutely possible on a PC with a decent GPU or a cluster of them, depending on energy costs and how many coffees you can tolerate making or series episodes you can tolerate watching while you train, refine and wait for outputs.

Using big tech's compute and API resources would be the way to go. But in a way you'll be feeding the beast and this wild west period we've entered won't last. Regulations are inbound. With some entertaining events around it in the meantime. If we want to have our cake and eat it, we're going to need the whole semiconductor industry blown wide open, IP and licensing for ISA's and software dismantled, the works. Else the fruits of human innovation in these areas will continue to flow upstream and concentrate.

 

[Eth]

Let us know. It would be very helpful to know. I'm just thinking out loud. Probably build best of class reliability is tops hardware-wise. Opensource with a good team behind it, with good track record in other earlier projects, barebones functionality that suit minimum reqt's for satisfactory performance. It is assumed that the ability to input text information would come easy from the existing repositories (such as of @Eth 's). Text-based data would be enough to begin with.

Using mine for now would be sufficient, although once I get all the interviews transcribed then it would be as close as we could get.

Just to let y'all know, one of my pals over at Chadnet has already make a pseudo-peat chatbot, you can play around with it at peatbot.com. It's pretty fun to mess around with, of course it's not perfect because all it does is source the context it refers to from the small source of information in the articles, but it still does a somewhat decent job. I think he sourced the context from either mine or chadnet's directory (wouldn't make a big difference regardless). Like I said, once I get all the interviews transcribed then it will be much more accurate.

 

[ThinPicking]

Just to let y'all know, one of my pals over at Chadnet has already make a pseudo-peat chatbot, you can play around with it at peatbot.com. It's pretty fun to mess around with, of course it's not perfect because all it does is source the context it refers to from the small source of information in the articles, but it still does a somewhat decent job. I think he sourced the context from either mine or chadnet's directory (wouldn't make a big difference regardless). Like I said, once I get all the interviews transcribed then it will be much more accurate.

Very cool with some money to burn.

Google Colaboratory

colab.research.google.com

 

[ThinPicking]

When people are locked into their 15 minute smart cities, with their smart appliances and GMO foods, they will finally realise that their inability to oxidize glucose is political.

Oh damn. What a line.

 

[purple pill]

Using mine for now would be sufficient, although once I get all the interviews transcribed then it would be as close as we could get.

Just to let y'all know, one of my pals over at Chadnet has already make a pseudo-peat chatbot, you can play around with it at peatbot.com. It's pretty fun to mess around with, of course it's not perfect because all it does is source the context it refers to from the small source of information in the articles, but it still does a somewhat decent job. I think he sourced the context from either mine or chadnet's directory (wouldn't make a big difference regardless). Like I said, once I get all the interviews transcribed then it will be much more accurate.

Awesome, thanks for this! Really saves time scrolling through newsletters when youre just looking for certain info or reminders