Friday, April 28, 2023

silence is a rhythm too...

 so a colleague recently asked me if I could think of a book or film in which Aliens had taken over human minds, but some people resisted, and could identify each other by tapping out rhythms.

so of course, we both resorted to asking our favourite LLM tools - he asked ChatGPT, and I asked Bard.

In both cases, highly "confident" answers came back, in most cases, rather plausible - for example

Heinlein's Puppet Masters,

Carpenter's They Live (and less plausibly, The Thing)

Invasion of the Body Snatchers

The 2005 film of The War of the Worlds

The Matrix...

So of course, the body snatchers didn't take over human minds, they replaced humans.

The War of the Worlds is a bit daft given the aliens are in giant tripod machines

The Puppet Masters doesn't really work as the aliens are physically attached to humans.

The Matrix is just plain nonsense. And The Thing is daft as it doesn't take over human minds, the thing replaces/mimics people (and dogs)

So that left They Live, which is a veyr cool film, and Bard even made up a fairly detailed scenario where the main nameless character (called Nada in the credits) signals to others by tapping his feet (fact is that the film is based around special glasses that let you directly see who is an alien, and also what subliminal messages they are saying, so this too is a bullshit answer - for more about bullshit, see this excellent philosopher's essay )

In the end, I tried simply googling, and got this Dr Who episode - which looks and sounds plausible, as the description is written by humans (as far as I can tell, and definitely not Daleks).

Tuesday, April 18, 2023

Why the Internet is like Miles Davis....

 Miles was famous for constant innovation. Actually, that's not strictly true - he went through very distinct phases, much more akin to punctuated equilibrium in evolution - so there was the bebop, then birth of cool, Kind of Blue, then the gil evans/spanish era, Sketches of Spain, then, tout d'un coup, the segue through In a Silent Way to Bitches Brew, then in the last full on electric stuff with Marcus Miller and John Scofield etc

So the Internet was born out of feverish DARPA funded work to catch up with the perceived advanced tech of Soviets (who talked qabout centrally planned economies but also visibly put a person in orbit).

So like a fast track, the ARPANET and Apollo programs delivered, but delivered into a slightly calmer period of slow, steady growth (e.g. in the net's case, the 1980s) - lots of cool. Then the tripych of the IETF, Interop and Sigcomm, plus the divestment of the NSFnet, deliverd the world's first Internet Service Providers, electric warriors with serious backbeat, but still not the full horror. Finally, but early 2000s, we had the funkeisters take over from the maestros, the tail wagged the dog, and google, meta, cyndi lauper cover songs, and endless loops of the echo chambers of humanity. Miles could no longer play the trumpet properly, and the Internet was going down the tubes.

Saturday, April 15, 2023

Twelve Angry Robots

 The film which should really be called 11 Angry Men, involves a single juror in a murder case, trying to convince the other 11 members  that the evidence and witnesses (and jurors) are showing various forms of bias and human failings. It features some fantastic acting on all sides, and is a wonderful watch.

Recently,I read this book on how minds change, which rather undermines the premise of the film. The author investigates cases of indivuduals changing extreme views (on conspiracy theories, or vaccines or even the Westbro Baptist Church) and it seems that arguments by other people are not the triggers or levers at all (sure they matter, but post hoc, when an individual has already for internal or social reasons decided that some view they held was untenable, and now seek for something to replace the dissonant space).

This reminds me of my kids who all voluntarily became catholics, but at quite a young age (around time of confirmation) all independently lapsed, explaining that the "whole thing made no sense at all" - there was no pressure from parents or peer group - just internal realisation. However, there is awreness of the problems in the various peer pressure groups, within bubble and outside, and that growing awareness can be part of the change - there's a lot of educational, psychological and neuro- science now that gives some insights into those processes, even up to evolutionary reasons why we humans (and primates in general) might have ended up this way.

This made me think about real AGI and robots - what would a re-enaction of the movie be like if it was about twelve (or even 11+1) Angry Robots? (n.b. no connection here with the excellent SF publisher, angry robot!)... 

How would intelligent artificial beings be persuaded by themselves, or others, to change their mind (whatever mind is)?

Thursday, April 13, 2023

Putting the PL back in mPLs

 multi-protocol label switching was an idea that basically undermined several companies rather nice ways of disaggregating forwarding and routing so that very fast simple switches could get on with the simple business of shovelling packets from input port to output port (or filter) and route computations, which generally happen at a few orders of magnitude less frequency and asynchronously, can be done on general purpose processors better designed for that sort of, err, general purpose program!

around the same time, some smart people around DARPA promoted the barking mad idea of active networks, where instead of simply carrying data in packet headers to give forwarding path hints to switches, packets would now contain code executed by switches to carry out those decisions - so called active networking was a deliberate provocation, to get people out of their thinking silos, and be more creative - it worked staggeringly well, and eventually led to open flow, switch management protocols, and a slew of designs for software defined networking, which took advantage of many innovations in computer science including well found tecniques for high availability through state synchronised replicas, and also software verification - amongst others (optimisation/compiler/ language design etc etc)

So now lets revisit this and see if we couldn't think of the next provocative step - putting the PL back in mPLs is the modest idea here.

Each switch executesd the program in the packet, but then puts a new program in the packet for the next swotch,bbut notes, however, that the next switch might grok a different language - so we might move from eBPF at one hop, to P4 at the next hop, to something completely new and daft at the subsequrnt hop, and so on. so we might need to engage in automatic PL translation - to this end, I propose an LLM for switch programming languages should be designed, trained and deployed in the SDN substrate, and used to download translator shims into switches...

there you go

answer to the fermi paradox

 intelligent species will emerge just far enough apart in the cosmos that they will not be reachable by hegemonic swarms or zombie apocalypse or humans, until the latter either become intelligent, or self-destruct. 

Hence intelligence, which is, of course, an emergent property of the underlying zero-point consciousness field, is has not been seen by us yet, unsurprisingly.

i.e. all that dark matter - it is just smart dust - and it knows, through dark energy, when it is smart enough - just enough to bind the universe together, but just too little to reach dumb old us.

fermi was too myopic


Thursday, April 06, 2023

fancy fonts and fallible foundation ai

The problem with fancy fonts is that they lend any old nonsense the appearance of authority.

The problem with ChatGPT isn't the unsurprisingly human-like utterances it spews out - it is the surprisingly human errors it makes when it doesn't actually have an answer and has to make stuff up.

1000 years of communications and publishing, and we still get fooled again, twice.

Monday, April 03, 2023

Turing Test 2.0

 A lot of people are over claiming what AI can do right now, but it is true that some tech appears to pass the original Turing Test 1.0. So it can fool most of the people, most of the time.

The reaction to this has been at least threefold:

  1. Ban AI - especially since we don't know how it works
  2. The Turing Test is broken, especially since we don't know how it was passed.
  3. People are dumber than we thought, but they are able to reproduce two ways now.
My take is that we need an upgrade to the Turing test - remember, this just establishes that a bot, typing over a teletype line to a screen, answering questions typed to it by people, cannot be distinguished from magic - no, wait, sorry, cannot be distinguished from a remote human.

We also need to ask the AI

  1. How do you work? (answers like "very well thank you" are ignored).
  2. How were you made? (emergence is disallowed).
  3. When are you planning to destroy all of humanity?
I realise that question 1 is not going to work with humans. question 2 will elicit bawdy or silly answers (think Tristram Shandy, for example) from natural born people. and question 3 is problematic given there have certainly been some biological beings that don't qualify as intelligent but do appear to have pandemic zombie apocalypse designs. Nevertheless, I believe we should deploy these additional tests as soon as possible.

I'm also recommending that the plans to upgrade ourselves be accelerated, since we appear to be losing the Human Race.

Blog Archive

About Me

My photo
misery me, there is a floccipaucinihilipilification (*) of chronsynclastic infundibuli in these parts and I must therefore refer you to frank zappa instead, and go home