Ironin i Torys förhållande till Brexit

Torys gick ut det Brittiska valet som storsegrare med löfte om en Brexit. Men man måste ändå tycka det är lite ironiskt att Engelsmännen vill gå ur ett förbund (EU) men tvinga Skottarna att stanna kvar i sitt (GB).

Men världen själv, varat omkring oss och inom oss, är aldrig ensidig, därom måste man tiga

Läste precis ut Hermann Hesses fantastiska Sidhartha. Sist jag läste den vad jag runt 16, så kändes som det var dags igen. Det är verkligen en fin bok. En insikt som kom till mej var att Sidhartha når samma slutsats som Wittgenstein. Varom man inte kan tala, därom måste man tiga. Vi kan bara säga så mycket om världen - varat omkring oss är aldrig ensidigt. Och därmed kommer vi aldrig kunna säga allt. Vi kommer alltid behöva ett förfinat språk och taget till sin extrem kommer vi aldrig kunna uttrycka det mest fundamentala i vår verklighet. Språket är i varat. Därmed kan språket aldrig komma ut varat, komma bortom och helt förklara det.
Därför behöver vi modeller. Dessa är vad vi kan prata om. Dessa är vad vi kan veta något om. Dessa är vad som kan vara användbart. Resten måste vi tiga om.

Seeing reality as it is is still a possibility. In fact, that is what we do. More or less

Hoffman just published another paper discussing his computer simulations indicating it is better to not experience reality as it is.

The ideas as stated are of course not new. They are present already in Plato's cave myth. But what should be pointed out I think, is that the ideas are actually not very interesting. Or perhaps even scientifically wrong.

First. We describe reality with models that approximate it. Hence, there is nothing new to anyone that we do not experience reality exactly as is. But what is more important is that there is, taking the scientific process to its limit over time, really nothing else to say about reality. Per definition. If there had been, it would be covered by some scientific theory. This is kind of the argument also against a supernatural being which I yet have not met any theist having a reasonable argument against: if an entity can interact with reality, that mechanism of interaction can be probed, and thus modeled. And then it is not supernatural any more. If there is no such interaction mechanism, they the entity has no implication on our reality.
Thus, in a proof by contradiction, there cannot be any supernatural entity.

On another note, of course an organism can be greatly successful and still not perceive reality fully. That's trivial. Take treas or any other plant. Surely their perceptive abilities are very limited and do not cover the whole of reality. They are still very successful as a spices. There is no problem here. However, a tree does also create models of the world. It is about light and temperature and nutrition. And the trees act according to those models. Surely, those models are quite good, they have made trees very successful. They approximate reality and the approximation is in direct relation to what that species needs. Trees cannot move, and hence they do not need models for kinematics.

This actually leads to an interesting thought - is evolution creating model-generating organisms? It looks like it. The genes are really an information processing system and as such, they are operating on some model assumption.

Seriously California, you are not making you best effort!

I recently had the opportunity to visit Silicon Valley. The stay was a pleasure, interesting meetings and great weather. But I must say I got quite upset. It was my first visit there, and it was very clear - California: you are not making your best effort here. I am of course talking about climate. Some examples:

• San Jose is, I don't know how, large. 4-5 lane roads, enormous parking lots and virtually no public transportation. Only 1-store buildings. This is not a building culture that is sustainable.
• On every single trash can I could find, it was printed "land fill". What the heck California? This is something we stopped doing in Sweden 40-50 years ago.
• In my hotel the cleaning lady changed my toilet roll even though it was at least 1/3 of the paper left. Enough to last several days for me! Just because it looks good. I am sure they throw that roll.

These are just some observations of the fact that this "liberal" state does not care. Here, in the richest area in the wold - without competition - nobody really cares about the environment. They care about exits, autonomous cars and apps.
I think California should feel ashamed. You have been the most lucky, yes lucky, in the world and you are not prepared to give anything back. That's BS. Shame on you!

Regeringens symbolism

Följande foto är från regeringens budgetpresentation ht 2019:

Här kan man se att både Stefan och Isabella väljer att bära sina partisymboler på slaget. Är inte detta underligt för en regering? Borde det inte vara svenska flaggan om någon de ska bära? Vilken är deras främsta prioritet egentligen? Sverige eller partiet?

What's it like to be a human

There are of course many ideas about this. So here is one at least I have not heard of before.

The unique thing with being human is that we can form models. The rest of us is really just ordinary animals. But the capability to form and reason in models makes us unique and exceptional.

Why? Well, the idea is that models are extremely powerful tools for comprehending the world and creating things. All achievements we have accomplished during the last 2000 years have been through models actually.

Further, to perform any task that is a bit more complex, it require that there is a model in its basis.

Some things we can do that do not require a model basis is
- creating art
- playing chess
These would seem to be unique human capabilities. But I would argue they are a side effect of our large brain. In principle a monkey could do it. However, a monkey could never build a rocket. Why? Because that require massive use of models to get everything right. And they can't do that.

The fear of AI is completely misleading

Many people are raising a fear for AI and that a super intelligence might not be aligned with the interests of humans. There are many reasons to be skeptical against such a view. Some have been covered in this blog before. Here is another one.

Say that is turns out artificial super intelligence is the outcome of the processes we see now. Say that such ASI outperform humans in most tasks. Why is it then a bad thing that they supersede us? If they are the next step in evolution, is it not a good thing that it materializes?

Very many of the arguments against ASI are really grounded in fear. But fear is not a good basis for making decisions.

About AGI

Alison Gopnik pointed out something very important about AGI and general capabilities. As you increase certain capabilities, e.i. you strengthen the priors, you at the same time reduce a capability of generalizing.

This could be used to derive some theorem about limits of capability. Widening it somewhere will be necessity narrow it down somewhere else.

So creating a general intelligence like a human will forcibly reduce some other capabilities, e.g. in playing chess.

This is an indication of that there is a limit in how intelligent an entity can be and also that AGI might not supersede us after all.

Integrated functional co-evolution

This might be a completely well understood and known thing, but I just realized that there is co-evolution within a single organism. Easiest is probably to illustrate with an example. Why do humans have arms and hands? They are really not very impressive compared to e.g. a cat's legs - sharp claws and strong so that they can jump 4-5 times their own height.
But the hands and arms have a completely unique feature - they are very generic. We can do many things with them. This would not be a very useful feature for a cat since they are fairly dump, and would thus not know how to use these generic tools. So, to be useful, they also have to go with a very large brain. Just as it happens, humans do have large brains. This is not a coincidence. Our brains have evolved together with our generic arms and hands. As we became smarter, new advanced things could be made with our hands, and then it would be beneficial to have more generic and even better hands and arms. E.g. with finer resolution.

This kind of integrated functional co-evolution is a different kind of evolution compared to how spices evolve in relation to it's surrounding. It is also different compared to ordinary co-evolution where different organisms co-evolve so e.g. be more and more specialized.

It would be interesting to hear if anyone can tell if this is well know process, or something novel. It would also be nice to test this in some kind to computer simulation.

Naturvetarföraktet måste stävjas

Tittade på SVTs Bäst i test med sonen ikväll. Dom hade med en uppgift där deltagarna skulle lista ut koden till ett lås. Svaret var 314 och på flera ställen var ledtråden π. Det är såklart kul och nästan rätt, bara fel på tusendelarna och framåt i decimalutvecklingen.

Men sedan tog kreativiteten fart och en ledtråd var

exp(i)*x +1= 0
x=?

Det är såklart kanonkul att SVT vill ha med en så berömt ekvation i ett populärt TV-program. Men det blir inte roligt när den är skriven som ovan, dvs fel. Detta fostrar en kultur kring naturvetenskap som obegripligt och mystiskt som inte är ok.

SVT och andra offentliga opinionsbildare har ett ansvar, Sveriges ingenjörer, naturvetare och matematiker är instrumentala i att skapa det välstånd vi lever i. Vi förtjänar att tas på större allvar och fler bör inspireras att utbilda sig inom dessa ämnen. Då kan man inte driva med det på detta sättet.

Tyvärr är det inte första gången SVT gör på detta sättet. I programmet På spåret, där frågorna typiskt är så omöjliga att ingen mer än en journalist kan besvara dom, är det igen helt ok att så fort det kommer en fråga med minsta naturvetenskaplig koppling vara helt Rudis.

SVT ryck upp dej. Vi som byggt landets välstånd förtjänas att tas på större allvar.

Carl Schlyter borde inte startat ett nytt parti utan...

... istället borde han gjort gemensam sak med FI. Dessa partier, drivna av specifika och viktiga samhällsproblem, har allt att vinna på att gå samman. Det är säkerligen så att nästan alla FIs väljare kan ställa upp också på en tydligare miljöpolitik. Och med de starka profiler som Carl och FIs ledare har skulle de tillsammans kunna skapa något större än vad de kan klara själva. Och Sverige behöver faktiskt inte fler partier. Vi har redan för många.

Peter Englund för ett mycket underligt resonemang

Läste om Reformisterna inom Socialdemokraterna som bl.a vill ta fram ett system där Riksbanken ska kunna tillhandahålla tjänster som idag tillhandahålls av kommersiella banker.
Professor Peter Englund kritiserar förslaget och säger att det ibland uppstår förtroendeposter för kommersiella banker och det i säkert sådant läge riskerar slå ut dessa om det finns ett statligt alternativ.
Detta resonemang är ganska absurt som kritik av förslaget. Om de kommersiella bankerna sätter sig i sådana förtroendekriser att privata sparare inte längre vill ha kvar sitt kapital hos dem så kan det väl inte lösningen vara att begränsa medborgares frihet och tvinga dem ha kvar sina pengar där för att bankerna ska överleva. Det blir fullständigt befängt. Det är väl snarare i så fall ett argument för att de kommersiella bankerna fått alldeles för stort svängrum under alldeles för lång tid.
Det är tråkigt när akademiker låter ideologi gå före logik. Det urholkar förtroendet för de ekonomiska vetenskaperna. Tänk om Peter Englund.

On free will

I just read the first few chapter in Sam Harris excellent Free Will. He is of course correct. Free will is (almost certainly) an illusion.

However. He is sort of pointing to something in the book which I have been thinking about before and which is more or less captured by the Turing test. A device has free will if its actions cannot be foreseen and they are known to be at least partially causal.

This connects to a formulation I made a long time ago about free will. A system has free will if it is by it self sufficient to determine some action outcome given some stimulus. This is an attempt to differentiate it from stochastic and deterministic systems. In this sense, a system possessing memory or hysteresis has free will.

On consciousness (again)

I just listened to Sams pod with David Chalmers on consciousness. They get it almost right about 50 minutes into the show when they talk about panpsychism and the ideas that everything is conscious at some level. What they miss is that they should really (as I have done in a previous blog p lost) define consciousness as interaction. They talk about information processing which is a complicated non-physical (but rather mathematical) concept. The correct concept to use is interaction which is very physical. And with that, everything is conscious. But at different levels.

These ideas also explain in a natural why we as humans cannot grasp how a chair or even a worm can be conscious. The problem is that the kind of interaction these entities takes part of is so vastly different from ours.

In this sense, consciousness is an illusion. It is a local phenomenon in an hierarchical structure of more and more complex interactions. And a consciousness operating as a certain level and with certain complex interactions cannot grasp, or even experience, a consciousness at a vastly different consciousness level.

It is the self-experience of the interactions occurring in a particular physical structure. We as humans can experience cars and chairs. Complex objects and we interact with these. (This is also probably why small children and illiterates attribute consciousness to objects that are not conscious at our level - it is the only thing we understand)

Later in the show Sam and David got into the thoughts on simulated worlds and that perhaps we live in one. This is also in some sense obvious true as I have also discussed before. We have some experience. That is mediated through our senses. But the world in it self (in Kantian terms) is something we can have no real idea about. Thus we do live in a simulation. It is impossible to say something outside of our experience of the world about the world.

We will not create true AI without understanding it

Everyone Is talking about AI. How fantastic it will be or how dystopic it will be. How we will create a super intelligence without understanding it. How it might pose a threat or how we will transfer our consciousness to some non-biological hardware.

At the same time a picture is drawn that this might happened with us humans in the back seat. Without us understanding it happen, or the consequences.

This is all ridiculous. Is there a single technical achievement in history that has just happened? The closest is perhaps the discovery of penicillin.

But in general every progress is preceded by careful analysis and hard work. Often by a large group of humans. We will not wake up one day and AI is just here, ruling the world. We will have to understand self-awareness, free will and consciousness long before we will be able to create it. There is no doubt about it. Anyone saying anything else is a believer of magic and fantasy.

And as for consciousness, self-awareness and free will, we haven't got the slightest idea what these things are. There are some fuzzy sociological ideas, mirror-experiments on chimps and the notion that we have to have free will because we have no other choice (not to mention the out-right ridiculous arguments put forward by religious proponents about God giving us free will and other nonsense crap). But these are all non-constructive. Useless. We will not be able to create true AI before we understand it. What people perceive as AI today is simply complex algorithms running on very fast computation hardware. Hence it is just Moore's law still in operation. Nothing else.

I am not saying that Moore's law won't be the thing that actually gives us AI, that it is just emergence. That might well be; but we will understand it before we create it.