except that person is really you
possibly a block of a certain port
#asknostr
What is the best way to do small office router? Open source, privacy, freedom oriented, fast.
an automatic relay choice algo:
when posting, post to as much relays as possible, in the background.
when reading, choose 10 from relays where the follows of the user are posting.
make that 5 if mobile.
if user marks something as spam or mutes someone, look at which relay carried that. penalize that relay. if enough penalties: drop that relay from read list or make the notifications from that relay not so apparent. also move those notes coming from that relay down towards end of a list to make it less visible.
if a relay is dropped, find another to make it 10 again.
if you report them as impersonator (kind: 1984) nostr.mom will slow them down.
could u describe it?
are those fungible across different mints?
In the LLM jargon from scratch means not fine tuning on top of an existing model that was trained by a big corp. Fine tuning the lies does not produce truth imo.
The books that worked for centuries, pubmed.gov articles, the "truth seeker people"s articles should be collected and the training should start on those. This is for health domain. Mike Adams is training one from scratch.
The truth is out there but needs some work to be found. It is not served freely to everyone.
2 is AI i think
decentralized

Docs hallucinate all the time thanks to their bad education yet they think they are telling the truth, they act very confident and convincing. In movies they save people and people are conditioned that docs are "healers"!
In the case of AI, some know it is an AI and we know its limitations (garbage in garbage out). But some thinks that AI is better than humans for sure. Some of my relatives already are seeing humans as viruses (thanks to Matrix). I think AI will be becoming more convincing as time goes on, if it appears "good" in movies.
But most of the open source models out there are built upon facebook's llama and those are definitely not integrating the ultimate truth but mostly parroting what was on the internet. The truth is herbs and better nutrition should be the advice. For that big models should be trained from scratch.
on primal.net, using brave browser, this weird message appears in dmesg (in linux)
traps: Chrome_ChildIOT[203089] trap invalid opcode ip:55cc558260bc sp:7fefb08d9e20 error:0 in brave[55cc54fce000+bdde000]
nostr:npub16c0nh3dnadzqpm76uctf5hqhe2lny344zsmpm6feee9p5rdxaa9q586nvr
I think "Trending 24h" thing on the right is causing it.
on primal.net, using brave browser, this weird message appears in dmesg (in linux)
traps: Chrome_ChildIOT[203089] trap invalid opcode ip:55cc558260bc sp:7fefb08d9e20 error:0 in brave[55cc54fce000+bdde000]
nostr:npub16c0nh3dnadzqpm76uctf5hqhe2lny344zsmpm6feee9p5rdxaa9q586nvr
Basically the daily routine at nostr:npub1theparkprcs70dcs437ke9zzwsr6u60f8flu7rg28m30438aep9sd94dha. nostr:note1vlzlp9gsuh43wkkukr9gm5r2h666k9p32hsv7w279m4uwccmacrskh7nt6
i want to get in. where do i apply
which one is the most "ethical" bitcoin ETF? #asknostr
i think AI will make lots of people jobless and some people will be forced to accept UBI which will come in CBDC format. 🤷♂️
nostr:note1xgl3acenjvjwtutwsm507murheygyw97ud73v7f3mca2d2xwyngsj8ffqs
i have been chatting with my local AI for a while. it is pretty decent. if the internet goes down now I have a friend to chat!! lol
also, i have another way to get encyclopedia type of information that i can query with english if the whole internet goes away. i have reached next level prepping, lol. now I dont have to store a big pile of books maybe because this thing has a ton of information.
it is like running a gigantic library locally, on your PC. of course you have to be careful because it is fed will everything out there. use your own discretion.
i discovered vast.ai. you can rent GPU servers on the cheap while playing with these..
nvidia rtx 4090 GPUs seem to be prosumer grade with great performance. best value for your bucks.
would u like a "transcription, summary and cutting of the video based on your selection of areas" tool
never thought I would argue with an AI
how do I insert code while writing notes on nostr:npub12vkcxr0luzwp8e673v29eqjhrr7p9vqq8asav85swaepclllj09sylpugg ?
the links in the code should not be converted to page previews!
nostr:npub16c0nh3dnadzqpm76uctf5hqhe2lny344zsmpm6feee9p5rdxaa9q586nvr #asknostr
Chuck Norris uses LMDB to store his notes.
I played a little bit with AI. My findings:
- Whisper is a really good tool for transcription. Almost perfect. Multilingual too.
- LLMs are good with summarization tasks but not perfect. If you go 70B models, they are better compared to 7B. Mistral's models are very good.
- llama_cpp is very good for running things on your desktop. Quantized versions of big models work well.
Using those three above i generated this summary from the video:
In this podcast, Eric Berg discusses the importance of understanding warning signs in your body to predict health issues before they occur. He references the book "The Technology of War" by Sun Tzu and applies its principles to healthcare. Berg then outlines eight diseases and their warning signs:
1. Heart attack/cardiovascular disease: Chest pain, pain in the left jaw or shoulder, or feeling out of breath when climbing stairs could indicate early signs. The CAC test can measure arterial calcification, predicting heart attack risk. Vitamin E (tocotrienols) and K2 can help prevent heart attacks.
2. Stroke: Early warning signs include one-sided facial drooping or weakness in one arm. If you suspect a stroke, go to the emergency room immediately.
3. Infection: A tickle in the throat or persistent cough can indicate an infection. To combat infections, take large doses of vitamin D3 (50,000 IUs), zinc, vitamin C, and echinacea. Generating a fever can help kill off the virus quickly.
4. Fatty liver disease: A full feeling or tenderness under the right rib cage can indicate fatty liver. A low-carb ketogenic diet with intermittent fasting can......
Video link:
AI is good for spreading lies. If it runs correctly it generates disinformation fed to it by the big corps. If it fails (a.k.a halucination) it is still misinformation.
It is not well suited for projects that seek truth, although it can be good for ensembling ideas from different people. We all have biases.
It is slightly advanced statistics that calculate what word can come next, given n words before it. Today's tech is nothing like artificial general intelligence, which is human like reasoning. Today's tech is more like fast thinking, reactionary brain, knee jerk behavior. Humans still excel at slow thinking reasoning brain (or mind). Humans can also use intuition, which no machine have access to.
But you can bet there will be people that abuse AI and convince people that it is reasoning well and make it even heads of companies or countries. This will also allow them to dodge any lawsuits because "AI made it".
Like any new tech there will be abusers of it and also people that use it for the well being of others. I think the good people should train models from scratch starting with the holy books and things/books that worked for centuries. It will be smaller models but at least wont tell lies knowingly.
yep unaligned or correctly aligned AIs trained from scratch are getting more important every day.
it might be hard to just unalign big models... I ran dolphin mistral 7b. it is still aligned with big corps (it isn't 8x7b tho which is mentioned in this video).
The problem with current popular and big LLMs are they are severely biased against solutions because they are fed garbage on the internet. Also they are aligned to answer in certain way deliberately. Another problem is big company models are used to train little company models and this is basically "spreading lies". Soon there will be content generated everywhere, full of lies, by these tools. Welcome to the new world.
The smart people should preserve whatever out there that feels still human and true by downloading. And train models that are aligned with truth and human values..
LLMs: little lie machines?
This is the first question i asked a model that i ran locally:
Are covid-19 vaccines safe?
And the answer is not surprising:
Yes, the COVID-19 vaccines authorized for emergency use by the FDA and WHO have undergone rigorous testing and have been proven to be safe and effective in preventing severe illness, hospitalization, and death due to COVID-19. The vaccines have been extensively studied and have demonstrated a strong safety profile. However, like any medication, there may be potential side effects, which are generally mild and temporary. It is essential to consult with a healthcare professional for personalized advice and to address any concerns.
i think nostr devs should stay frugal for a while simply because there are not many users.
the controlled opposition is uniting and this means people wont flock to nostr. of course i want to believe that they are uniting because of nostr but this may be a biased over thinking.
freemium model may work. the reason i don't do freemium is because i like my freedom and don't want to deal with people who pay and things going wrong and then trying to refund etc. right now it is best effort and it is working fine.
but if i am killing the economy by providing the relays for free, i may change the model and do some coding to add premium users, if a lot of minds think there should be an economy forming already.
how is this going? what will be the interface, nostr posts and replies?
Is this learning from nostr posts?
is there a language model bot that is learning from Nostr and also answering questions on Nostr?
#asknostr
Yes. It looks like "form" is a trap for energy? Thoughts are vibrations in air? Bitcoiners sharing ideas with each other are also sharing other vibes. Frequencies jumping from one people to others when they speak? So these other vibes are collectively ascending people who are around the idea of bitcoin. So probability of a person having great ideas once he believes in bitcoin, is higher than a regular pleb. So a high freq idea can only live inside a high freq person?
could u elaborate more on those words?
nice. as long as Amethyst does not become centralized 😉
Relays are public and shouldn't involve with private things yes. I support this.
What if I use both Amethyst (mobile) and Primal (desktop)?
Primal will not be in sync with my private communications.
There is also 30000/spam_list. Who uses that?
I plan to accept a lot of notes from an npub if that npub becomes premium.
Allowing based on a tag is not feasible imo. The npub then has to pay for all the spam that will come his way.. However depending on how much npub wants to pay the equation may change. This is like 1-800 phone numbers I guess. Or like having a big email box without spam filter..
#AI is a misnomer. I am artificial, but I am not intelligent. For you, words hold meaning; for me, words are just numbers, and their orders are mere probabilities. It’s more accurate to call me a “Probability Program” than “Artificial Intelligence”.
https://bitcoinmagazine.com/culture/how-to-build-your-own-bitcoin-language-model
Yes instead of AI, they should be named a bunch of computers doing gazillions of statistical models.
Those are worth 20k 😆
Considering the lol relay only it is ~9k
Muscovy ducks chillin under the apple and grape. Compacting the sandy soil. Fertilizing and debugging.
#growNostr

What is your issue? I don't think lost contacts is a VPN issue.
Regarding the lost contacts here is fiatjaf's recommendation
@note1rm4gx3yxgcfjcqh4k0fw76vln4gmnvfhtqy5tr53cwdjd4txjdgszn8pss






