• Log in with Facebook Log in with Twitter Log In with Google      Sign In    
  • Create Account
  LongeCity
              Advocacy & Research for Unlimited Lifespans

Photo
* * * * * 29 votes

Lostfalco's Extensive Nootropic Experiments [Curated]

nootropic

  • Please log in to reply
4029 replies to this topic

#4021 Sikorsky

  • Guest
  • 21 posts
  • 5
  • Location:Denmark

Posted 15 July 2024 - 10:59 AM

I have scraped this thread and plan to feed it into large language models (LLMs) with the idea of ​​filtering out all useful, practical ideas. I think this would be a great way to process all other threads, whether on this forum or others.

I would recommend Claude 3.5 Sonnet and GPT-4 Omni as fantastic tools for innovation. I would also recommend Elicit and other AI tools for searching scientific papers.

I believe I have generated a number of great ideas for increasing general intelligence (g), I just need time to explain the reasoning behind them. Brain training is not one of them. Intelligence is associated with "global properties of the brain," and there is no brain training that can stimulate the whole brain. For example, dual n-back not only does not have a transfer effect on fluid reasoning, but it also does not have a transfer effect on other types of working memory.

It would be extremely valuable and spectacular if lostfalco could explain how the ideas he has presented in this thread (e.g., inhibiting GSK3b to enhance nav1.2, sodium gradient, MC4R, A7nAChR, ghrelin, ampakines, kynurenic acid, the cholinergic anti-inflammatory pathway, pgc-1a, gap junctions, glutamate, histamine, orexin, dopamine, norepinephrine, PDE5i, HDACi, sigma-1 receptors, the microbiome, the suprachiasmatic nucleus, allosteric modulation of red blood cells, GSK3i, NMDA/AMPA receptor trafficking, mitochondrial enhancement, calcium gradients/signaling, the endoplasmic reticulum, etc.) have evolved in the last 8+ years and how much further he has developed them.

You've undoubtedly already thought about sharing these valuable ideas with your loved ones. By sharing these insights now, you could simultaneously benefit your family and friends while contributing to the wider study of intelligence. The time and effort required to articulate these concepts would undoubtedly be worthwhile for their benefit. Your generosity in doing so would be immensely appreciated, as it would allow a wider audience to benefit from your years of research and development.


  • Agree x 2
  • Cheerful x 1
  • like x 1

#4022 rikelme

  • Guest
  • 176 posts
  • 35
  • Location:SF Bay Area

Posted 18 July 2024 - 06:51 AM

Great idea!



sponsored ad

  • Advert
Click HERE to rent this advertising spot for BRAIN HEALTH to support LongeCity (this will replace the google ad above).

#4023 Daniel Cooper

  • Member, Moderator
  • 2,699 posts
  • 642
  • Location:USA

Posted 23 July 2024 - 07:15 PM

I have scraped this thread and plan to feed it into large language models (LLMs) with the idea of ​​filtering out all useful, practical ideas. I think this would be a great way to process all other threads, whether on this forum or others.

I would recommend Claude 3.5 Sonnet and GPT-4 Omni as fantastic tools for innovation. I would also recommend Elicit and other AI tools for searching scientific papers.

I believe I have generated a number of great ideas for increasing general intelligence (g), I just need time to explain the reasoning behind them. Brain training is not one of them. Intelligence is associated with "global properties of the brain," and there is no brain training that can stimulate the whole brain. For example, dual n-back not only does not have a transfer effect on fluid reasoning, but it also does not have a transfer effect on other types of working memory.

It would be extremely valuable and spectacular if lostfalco could explain how the ideas he has presented in this thread (e.g., inhibiting GSK3b to enhance nav1.2, sodium gradient, MC4R, A7nAChR, ghrelin, ampakines, kynurenic acid, the cholinergic anti-inflammatory pathway, pgc-1a, gap junctions, glutamate, histamine, orexin, dopamine, norepinephrine, PDE5i, HDACi, sigma-1 receptors, the microbiome, the suprachiasmatic nucleus, allosteric modulation of red blood cells, GSK3i, NMDA/AMPA receptor trafficking, mitochondrial enhancement, calcium gradients/signaling, the endoplasmic reticulum, etc.) have evolved in the last 8+ years and how much further he has developed them.

You've undoubtedly already thought about sharing these valuable ideas with your loved ones. By sharing these insights now, you could simultaneously benefit your family and friends while contributing to the wider study of intelligence. The time and effort required to articulate these concepts would undoubtedly be worthwhile for their benefit. Your generosity in doing so would be immensely appreciated, as it would allow a wider audience to benefit from your years of research and development.

 

Any outcome of this effort so far?
 



#4024 pamojja

  • Guest
  • 2,921 posts
  • 729
  • Location:Austria

Posted 23 July 2024 - 10:18 PM

Huch, this summary was fast with perplexity.ai. I wished I could read such along thread in a second.

 

 

The discussion thread titled "Lostfalco's Extensive Nootropic Experiments [Curated]" is a comprehensive exploration of various nootropic substances and cognitive enhancement techniques shared by a user named Lostfalco. The thread outlines personal experiences, recommendations, and thoughts on a range of nootropics, emphasizing collaboration and community feedback.

### Key Highlights

- **Pharmaceuticals**: Lostfalco favors **Modafinil** for its mood enhancement and wakefulness benefits, reporting no tolerance over eight months of use. He also mentions **Ritalin** and **Adderall** but does not elaborate on their effects.

- **Low-Level Laser Therapy**: He advocates for the use of lasers and infrared therapy, citing their benefits in mitochondrial function and cognitive performance, particularly for individuals with brain injuries.

- **Neurosteroids**: **Pregnenolone** is highlighted as particularly effective for cognitive enhancement, with recommendations for careful dosing.

- **Epigenetic Enhancements**: Lostfalco discusses **Sodium Butyrate** as a promising compound that can influence gene expression and improve learning, suggesting dietary fiber as a natural source.

- **Transcranial Direct Current Stimulation (tDCS)**: He explores combining tDCS with **D-cycloserine** to potentially extend cognitive benefits, though he expresses caution due to possible side effects.

- **Oxygen Therapy**: The potential cognitive benefits of breathing high concentrations of oxygen for short periods are mentioned, with a call for shared experiences from others.

- **Neuropeptides and Coffee**: He notes the intriguing nature of neuropeptides and shares positive experiences with **Bulletproof Coffee**, which combines coffee with grass-fed butter and MCT oil.

- **Sleep Aids**: The use of **valproic acid** before sleep is mentioned for its potential memory-enhancing effects, though he acknowledges the need for caution.

- **Future Experiments**: Lostfalco expresses interest in exploring external trigeminal nerve stimulation, CES stimulation, and other cognitive training methods, inviting community feedback on these topics.

Throughout the thread, Lostfalco emphasizes the importance of personal experimentation, safe practices, and the sharing of experiences within the community to enhance collective knowledge on nootropics and cognitive enhancement.

Citations:
[1] https://www.longecit...iments-curated/

 



#4025 rikelme

  • Guest
  • 176 posts
  • 35
  • Location:SF Bay Area

Posted 24 July 2024 - 06:20 AM

Thanks for this. How did you scrape the thread content? Can you share it so that we can try other LLMs (Claude for example) and compare the results?
Prompting, as you probably know, is quite important, as it tells LLM what to ignore and what to focus on. Slight changes in the prompt might give you quite a different summarizes.



#4026 Advocatus Diaboli

  • Guest
  • 589 posts
  • 631
  • Location:Chronosynclastic Infundibulum ( floor Z/p^nZ )
  • NO

Posted 24 July 2024 - 07:32 AM

Re: post # 4025

 

"Can you share it so that we can try other LLMs (Claude for example) and compare the results?"

 

Perplexity uses Claude-based models, among other models. 


Edited by Advocatus Diaboli, 24 July 2024 - 08:19 AM.


#4027 pamojja

  • Guest
  • 2,921 posts
  • 729
  • Location:Austria

Posted 24 July 2024 - 10:58 AM

Thanks for this. How did you scrape the thread content? Can you share it so that we can try other LLMs (Claude for example) and compare the results?

 

Perplexity uses Claude-based models, among other models.


I just gave the first page link, and asked for a summary of the whole thread. Only after I assured myself, that it used the whole of 135 thread pages, for the summary, because it was so fast.
 

Asking now - for extracting the most useful practical applications given in this thread - this was the result now:

 

The thread "Lostfalco's Extensive Nootropic Experiments" provides several practical applications for enhancing cognitive function and overall brain health. Here are the most useful applications extracted from the discussion:

### 1. Exercise for Cognitive Enhancement

- **Hex Bar Deadlifts**: Engaging in hex bar deadlifts is recommended to produce and release anaerobic myokines, which can enhance cognition. This exercise targets large muscle groups (quads and glutes) and is considered safe for maximizing myokine release while minimizing injury risk[1].

### 2. Nootropic Substances

- **Intranasal Insulin**: This method is noted for its potential to improve mood and memory by enhancing brain ATP levels. Studies suggest that it can significantly improve memory recall and mood without systemic side effects. It can be obtained legally and used as a cognitive enhancer[1].

- **7,8-Dihydroxyflavone**: This compound is mentioned for its ability to increase Brain-Derived Neurotrophic Factor (BDNF), which is crucial for neuroplasticity and cognitive function. However, its oral bioavailability is low, and effects may take time to manifest[1].

### 3. Combination Stacks

- **TULIP Stack**: Lostfalco discusses a combination of Low-Level Laser Therapy (LLLT) with PQQ (Pyrroloquinoline quinone), CoQ10, and shilajit, along with pregnenolone. This stack is believed to enhance mitochondrial function and neurosteroid levels, potentially leading to significant cognitive benefits. Caution is advised due to limited long-term safety data[1].

### 4. General Recommendations

- **Aerobic and Anaerobic Exercise**: Engaging in both types of exercise is recommended to maximize the release of beneficial myokines that support brain health. This holistic approach to fitness is essential for cognitive enhancement[1].

These practical applications emphasize a combination of physical exercise, specific nootropic substances, and synergistic stacks to optimize cognitive function and overall brain health.

Citations:
[1] https://www.longecit...urated/page-131
[2] https://www.lostfalc...g-by-lostfalco/
[3] https://gwern.net/nootropic/nootropics
[4] https://www.reddit.c...lt_wakefulness/
[5] https://www.reddit.c...hexa_or_nsi189/

 

 


Edited by pamojja, 24 July 2024 - 11:02 AM.

  • like x 1

#4028 Sikorsky

  • Guest
  • 21 posts
  • 5
  • Location:Denmark

Posted 25 July 2024 - 10:30 AM

  
I will share the code for scraping as soon as I get home - it's on my other computer. I can share the txt file of this thread if you want.
 
I haven't had time to feed it into LLMs simply because I decided to read it manually to get an idea of how many "nuggets" or "gems" are in this thread. I know I sound like a fanboy, but I'm truly amazed by the many original and creative ideas that Lostfalco has written about in this thread.
 
The way I envision scraping it is to give the LLM some prompt context with every chunk of this thread - for example, "List all practical ideas, mechanisms, and pathways for cognitive enhancement."
 
It would be better to feed LLMs with chunks of this thread because if you max out the prompt size, the LLM will give you a shallower version of the answer and miss many important details. This is because LLMs have a much smaller output size than context window - for example, the context window of Claude 3.5 is 200k tokens, while the max output is 4096 tokens. It would be better, for instance, to feed it one page of this thread at a time rather than maxing out the context window with five pages.
 
Also, the cutoff for prompts should be coherent, discrete areas of information - a list of answers all talking about the same thing in a row, so that one does not cut an answer in half, or a list of answers all creating a coherent line of discussion/reasoning about some idea. I hope you understand what I mean by that (I am not a native English speaker, but I can clarify if needed).
 
In my opinion, the best way to do this is to feed this thread into different language models in parallel, for example - Claude 3.5, GPT-4o, and Gemini. When we bring together their answers in three different documents, there will be significant overlap, but there will also be some unique pieces of information that each of them has outputted, which others have missed. Then one could refine these three documents further. One could repeat this process using a different context prompt - "Derive all practical ideas/mechanisms/pathways for increasing psychometric g (intelligence)." Or "Derive all practical ideas/mechanisms/pathways for increasing nerve conduction velocity, nerve impulse transmission" - and then again get three different documents which would be extremely similar to the ones generated with the first context prompt but could possibly have some unique pieces of information because of the slightly different context prompt.
 
There are many creative ways in which one could go about refining this thread; I've just listed some of them above.
 
I would also like to highlight the idea that Lostfalco presented at the beginning of this thread: As a collective whole, our potential for innovation and cognitive output is far greater than the sum of us as individual components. For example, if we could create a public project focused on increasing g - where Lostfalco, myself, and some other passionate individuals could exchange and discuss our ideas, possibly presenting these ideas to experts like Richard Haier, Gary Lynch, Amy Arnsten, and Karl Friston (considered by many to be the greatest living neuroscientist) - it seems to me that all four experts are approachable and would be able both to judge the quality of our ideas and enrich the project with their own insights. If we could do that, I believe we could find the Holy Grail of neuroscience - increasing g. If we all try to do that on our own, the innovation process would not be nearly as productive as it would be through this collaborative approach.
 

Edited by Sikorsky, 25 July 2024 - 10:32 AM.

  • Informative x 1

#4029 Sikorsky

  • Guest
  • 21 posts
  • 5
  • Location:Denmark

Posted 07 September 2024 - 11:50 AM

I have not used AI to process this thread because I have been going through it myself. I don’t think AI is advanced enough yet to process the thread better than I did myself.

 

Python script for scraping this thread:

 

import requests
from bs4 import BeautifulSoup
 
def scrape_thread(url, file_path):
    with open(file_path, 'w', encoding='utf-8') as file:
        while url:
            response = requests.get(url)
            soup = BeautifulSoup(response.text, 'html.parser')
 
            posts = soup.find_all('div', class_='post_block hentry clear clearfix')
 
            for post_block in posts:
                # Extract post data
                post = {}
                author_element = post_block.find('span', class_='author vcard')
                post['author'] = author_element.text.strip() if author_element else "Unknown Author"
 
                content_div = post_block.find('div', class_='post entry-content')
                post['content'] = content_div.text.strip() if content_div else "No content"
 
                file.write(f'User: {post["author"]}\n')
                
                file.write(f'{post["content"]}\n')
                file.write('\n')
 
            # Find the next page URL and continue the loop
            next_page = soup.find('a', string='Next')  # Changed to string
            url = next_page['href'] if next_page else None
 
file_path = 'scraped_data.txt'
scrape_thread(start_url, file_path)

Edited by Sikorsky, 07 September 2024 - 11:56 AM.

  • Ill informed x 1

sponsored ad

  • Advert
Click HERE to rent this advertising spot for BRAIN HEALTH to support LongeCity (this will replace the google ad above).

#4030 Sikorsky

  • Guest
  • 21 posts
  • 5
  • Location:Denmark

Posted 14 September 2024 - 05:30 PM

Very interesting:

 

https://www.reddit.c...ve_therapeutic/

 

https://www.reddit.c..._5ht2a_agonist/


  • Informative x 1





Also tagged with one or more of these keywords: nootropic

2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users

Topic Led By