Internet requirements for human survival and issues beyond our solar system

 I have often urged us all never to forget the four biggest threats in the coming century to the very survival of the human species: (1) "AGI/IOT", actually the entire complex of instabilities emerging from the emerging new internet; (2) climate extinction; (3 and 4) misuse of biotech and nuclear tech. It would be unhealthy and biased to pay less attention to serious HOPE, the positive side, but to ignore the worst threats to our existence would be crazy.

The "AGI/IOT" threats are by far the worst, most central and most immediate. People talk about "threats to democracy," threats of war and autonomous weapons, threats of civil war 

(e.g.this morning, see https://www.npr.org/2022/01/11/1071082955/imagine-another-american-civil-war-but-this-time-in-every-state), about the consequences of unemployment >70%, possibilities for near-term economic collapse due to short circuits in the world economy, cyberwarfare, etc. These are all very large and very immediate problems, but they are all just symptoms of something even deeper and more stubborn. 

At http://www.werbos.com/How_to%20Build_Past_Emerging_Internet_Chaos.htm, I tried to explain 

the biggest, obvious linked problems, and the radical redesign of internet which would be necessary to avoid the terrible entropy threatening to engulf the world. 

BUT WILL IT HAPPEN? The modes of instability have become clearer to me this year, in part because I see how they are playing out before our eyes, but also because I have had a chance to understand the math better. Therefore, I will just copy over a summary I sent to a fellow expert yesterday, and a link to a directory which describes some of the (partial) fixes I understood by 2019:

===============================================

====================================================

If you could really DO that... a GENERAL platform to INTELLIGENTLY bring together the highest human and machine intelligence (WHICH REQUIRES NEW MACHINE INTELLIGENCE) , to be able to learn to collaborate in a general way without killing/undermining each other... THAT would fulfill the highest images Bill wrote down in that email. If that was not stated, or not clear to him, we have a communications problem.

But I see the design of such a platform (CUM ITS CORE PRIVILEGED APPS) as a huge research challenge, requiring a new network already of US communication (e.g. in a game which you and I and Bill play?), and drawing some of what I know of the AI and human requirements which no one else on earth knows yet.

My overview http://www.werbos.com/How_to%20Build_Past_Emerging_Internet_Chaos.htm
AND SUPPORTING FILES NOT YET OPEN solve some of the design problem, but this year I see further demands (modes of instability demanding firewalls)  I did not see before. In essence, firewalls like the Sarbanes Oxley rules needed to prevent collapse of electric power grid systems are necessary, BUT ARE SUSTAINABLE LONG-TERM only if they are implemented in privileged app programs compiled on hard cybersecure open systems, NOT overly dependent on human word inputs (with IOT direct sensing to robustify networks of truth).
Only a new caste system for apps can liberate humans from THEIR need for speciation and caste systems, a need intruding into political systems all over the world even now. Computers to liberate humans...

It also requires a new level of machine intelligence, outlined in new paper for an Elsevier journal, for which they promise the page proofs any day now.

======================================

https://drive.google.com/drive/folders/1cKKbdlmk2JVsD2sVfBKvcwcMEEwITvBz?usp=sharing

Just the tip of an iceberg...

The Elsevier paper gives a concrete path to a technology it calls QuATh, including application to being able to image what we see beyond our solar system with far more accuracy than any classical technologies, like what Milner's Breakthrough Initiatives are capable of. It is odd that Milner wants us to advertize our existence without knowing whom we might be advertizing TO.