Thanks for the update on Moore’s Law, I didn’t really look into it too much as most people when they write these articles concentrate on the same things I try to do something a little bit different

I see what you mean about consciousness intelligence mind and brain but if something like the singularity does occur, in my mind (or is it brain?) all of these four attributes will have occurred,

Moore’s Law (an empirical observation rather than an actual ‘law’ (in the scientific sense) actually states the number of transistors/level of integration/ how much you can stuff onto a sillicon chip doubles every 2 yrs or so. This doesn’t necessarily translate into a doubling of computer power, though to be fair you HAVE included a link to a more correct definition in the article.

The scecond point is rather trickier & has dogged many in this field continuously – the distinction between conciousness,intelligence, mind & brain. These terms tend to be used interchangeably, andd often quite sublty, making it difficult to pinpoint or describe which of these is actually being discussed.

Not a criticism per se of an otherwise fairly well-written article, but just illustrating the difficulties of writing about such n intricate subject.

I do like the idea though but perhaps that’s because I want to live forever. The singularity is almost turning into a religion for some people but I feel there are bits and pieces that are plausible and add to my view of the universe.

Yes, I also think that 40 years is wishful thinking but it is popular thinking and really we should be looking at 1000 to 1,000,000 years to get the singularity although we will develop up to that point and interface with machines to the point where the interface between humans and machines is completely seamless.

It would be nice if somebody could build a brain so I could transfer into it and see when the singularity does actually does occur.

My remaining problem is that I still do. There is no inherent qualitative change in a new technology, whether it is intelligence boosting (already happening, btw) or anything else. As Clarke roughly put it, any sufficiently advanced technology will appear magical (partly from unpredictability on longer periods). Without a qualitative change or any quantitative measure, it is still untestable woo.

But I’m sure Kurzweil is rich.

[Btw, as usual "artificial intelligence" research is here confused with actual artificial intelligence.

It is AFAIU already known (since 2003 IIRC) that large parts of the cortex is working with efficient coding. (To keep signals stable - on average one neuron/time step activated from one neuron signal.) This makes attempts of decoding dependent on the original network, i.e. will in practice be impossible for the considered time horizon. So no "merge" during the considered period, merely interfacing.

As the article itself makes clear by example, natural intelligence is embodied, so hardware dependent. Again an exceedingly difficult proposition to make a technological equivalent, and even more to emulate it fully in software. (If it is at all possible, some hardware is tough to emulate.)

The difficulty is on the order of figuring out how a genotype makes a phenotype. That has been unsolved in its totality for 100 years, and the details will keep scientists busy for another couple of decades. So the way to bet is IMHO that 40 years is clearly insufficient - especially as much fewer scientists works on this than on biology at large.]