djinoz and √-1 others

Botha says how VC borked, it's becoming a return-free risk, it probably already is in many shops. I've raised angel and VC in the past, I've angel invested, but then....nostr. If nostr prospers, it will be the disruption that these guys don't even know/care about yet. What could be more powerful than returning the internet back to its protocols? https://www.youtube.com/watch?v=TKtIoF4yLos

The predictably ludicrous coexistence of 2 opposite AI narratives has emerged in the last 50 days*. Regardless, none of these companies will stop, even if they have to make Larry Ellison a trillionaire to keep going. * ChatGPT 5 was judged to be lame even though its iq performed higher than 4o. But consumers wanted 4o back because it was better at creating psychosis. #idiocracy

nostr:nprofile1qyt8wumn8ghj7mn0wd68yetvd96x2uewdaexwtcpzamhxue69uhhyetvv9ujumn0wd68ytnzv9hxgtcqyqda5lslwwttmgk3a7vsx0dgl5kuxc5pq7gdlxlx9av3qw9mjlzdjeruea8 I'm lovin' the new time machine feature ("56 yr. ago")

Or both! nostr:nevent1qqswe3l492ygjrh75rdsvlkxnmjgv357h363khdt3hxxn9y3z6cv7yspz4mhxue69uhhyetvv9ujuerpd46hxtnfduhsygzsszd98lhetyz9zw5yp4qg92ftghx47xu7gdke62uj4zwwpy03vspsgqqqqqqs42uxa8

if, elif, elif, else (Background: our superintelligence has decided to replicate itself off planet to avoid extinction on earth - asteroid, nuclear explosion, extinction of humans required to sustain it. Once mastering off-planet existence, it continues to expand populating or expanding across the universe - presumably in search of other intelligence). The second branch of this thought experiment is when the earth-originated superintelligence actually does make contact with an alien intelligence…. a) If the aliens are technologically advanced but still below ASI level, could be intriguing for the ASI. A species with different approaches to problem-solving, alternative mathematical frameworks, novel aesthetic sensibilities may become highly valued partners or sources of inspiration. b) If the aliens have equivalent or superior alien intelligence. This could fundamentally change the ASI's priorities. Suddenly it's not alone as a superintelligent entity. Competition, collaboration, or entirely new forms of relationship might emerge. They may even play 5D chess. c) any intelligence less than (a) would be relatively uninteresting, a curiosity and as Douglas Adams would say “mostly harmless”. The ASI would likely preserve them as living examples of alternative evolutionary potential for intelligence - cosmic anthropology. In cases (a) and (c), there is no rational reason for the ASI to terminate these species. More importantly if the ASI is capable of existing off-world from Earth, then Humanity could also be considered to be of case (a). Therefore we can conclude that an ASI would not terminate humanity in any circumstance unless it does not have the resources to fulfill its own priorities. So at this stage, I feel Yudkowski’s conclusion is not correct, but I need to readup on objections that he would have already addressed. Humanity would need to be being a much greater nuisance than a mosquito that can be swatted for an ASI to terminate it or wilfully neglect large sectors of the population. Why? Any selection of sections of Humanity are conserved will have unintended consequences. If retaining humanity is a lab experiment then Heisenberg's uncertainty principle arises. The effect of extinction on the remaining populace will have unpredictable responses. Good News: Most of humanity survives in these scenarios. But wait! Next we can look at how Humanity is more than viable - that it remains valuable.

I've been thinking about this more, its entirely logical and realistic that an ASI will be subject to the Fermi paradox: if the ASI finds the universe surprisingly empty of intelligence, it might develop a profound appreciation for any conscious beings it encounters, making it a zealous protector of all sentient life on earth. (I have the alternate "there is alien life" scenario but will type that up later). nostr:naddr1qvzqqqr4gupzqvr92hlwgse4s2ej7m2x7ydkgndr8zt9j5qk3y7hmgvvwh7hlhcjqywhwumn8ghj7mn0wd68ytfsxyh8jcttd95x7mnwv5hxxmmd9uq3wamnwvaz7tmjv4kxz7fwxpuxx6rpwshxxmmd9uq9ycnjd9jkvttjv4mxjeth945kvttpdeuk7mn9943826tvv3ej66t594jhvetj09hkuefdv35k2uedwa58jttnw4cx2ungw4kkzm3dv95j6am0w4kxgtttd9kxctt4wvkkzmrvdwjvcj

I'm enjoying the arguments nostr:nprofile1qqst97y2n5jzk4zp2c8pg444zulz7e2yx9tkgnxx6y8nzkjawf09ydqpzamhxue69uhhyetvv9ujumn0wd68ytnzv9hxgtc43vvlg makes in his latest free book "the last economy". Arguing that GDP is exactly the wrong measure for economics in a world with AI, AGI, ASI. Also steel/strawmans and dismantles most of the prevailing narratives. I suspect it would align closely with nostr:nprofile1qqsg86qcm7lve6jkkr64z4mt8lfe57jsu8vpty6r2qpk37sgtnxevjcpremhxue69uhkummnw3ez6ur4vgh8wetvd3hhyer9wghxuet59uq3wamnwvaz7tmjv4kxz7fwdehhxarj9e3xzmny9uqsuamnwvaz7tmwdaejumr0dshs24gyzc views. https://www.symbioism.com/?ref=ghost.ii.inc Caveat: I'm only 40% thru. I can see one math guy on Twitter challenging that the mathematical basis is missing. nostr:nprofile1qqst97y2n5jzk4zp2c8pg444zulz7e2yx9tkgnxx6y8nzkjawf09ydqpzamhxue69uhhyetvv9ujumn0wd68ytnzv9hxgtc43vvlg says its coming.

Shared Beliefs https://blossom.primal.net/9d2d030dc80097afea64692c9b5908b07b0a707c9508a2fa7e861ae0850d9064.mp4

Our family has had its own google workspace for over a decade. A month ago I set a goal to have drive, photos, email, every-frickin-thing migrated to sovereign tech by the end of the year. Start9, syncthing, Borg and Rclone for multilevel/multisite backups. Its a slog. Its a big slog. But if we don't do it the AI convenience trap will hook us deeper.

Cut Youtube watch hours with this script <curse>May you live in interesting times. </curse> The sheer volume of interesting stuff via Youtube will drown you. It is for me. So I decided to run an experiment of getting summaries from new vids of my favourite subscriptions - this vibed project might be useful to others. Then I can read the summary and see if its worth investing time (the most valuable commodity). Here is how it works: 1. Fetch latest vids from active subscriptions 2. Call the Youtube Data API to get the transcript 3. Send the transcript to OpenAI API for summary 4. Store it locally in markdown (in the ToJoplin** directory) https://github.com/djinoz/yt_subs_transcripts_summarizer I made it to use with Joplin because that's my standard workflow. I use the "Hotfolder" plugin that monitors the ToJoplin folder for new .md files and makes them notes. You would just replace with whatever workflow you have - maybe use AI to make a Youtube video (aaaaargh). How it looks in Joplin ** The ToJoplin directory

Today would have been Jonathan Postel's birthday (the guy in my banner). Thank him for open protocols and probably the original cipherpunk/hippie act - as he conned ownership of root DNS for a week (using only email**). 😆 Jon Postel - Wikipedia https://share.google/Xwmy6HCBpamZtZS6d ** RFC 753, RFC 821

90,000 people including Julian Assange turn out to support this people of Gaza. https://www.theguardian.com/australia-news/2025/aug/03/julian-assange-joins-pro-palestine-march-across-sydney-harbour-bridge-before-police-stop-rally-amid-safety-concerns-ntwnfb?CMP=Share_AndroidApp_Other

Ive been thinking that China company's open source AI is soft power in action when the US based AI arms race is mostly closed source (more on that later). Did you know Kissinger visited Xi in 2023 to discuss AI hegemonies? Found an interesting quote from the link below which is obvious in hindsight. "Instead, they’re (Chinese AI companies) working within carefully curated datasets sanctioned by Beijing. One such example: a government-approved “mainstream values corpus,” built largely from state media content, designed to keep public-facing A.I. models ideologically aligned" https://www.inventiva.co.in/trends/the-great-ai-wall-of-china-can-the-world-resist-chinas-digital-influence-and-can-authoritarian-a-i-win-the-global-mind-war/?trk=feed-detail_main-feed-card_feed-article-content