Danny Hillis – “The Pattern on the Stone“
this is interesting when compared to how the machine learning ML (link!!!) things of recent not (ChatGPT, Dall-e?) how the lins that form inside the multipple layers of millions of connections cannot be explained by anyone; how the specific conencts form one layer to the next and how the weights of these connections are arrived att to get the results that are somewhat impressive (link to recnt breast cancer ML that allegedly does a way better job than hoomans)., in the enbd, we can’t know, can’t describe, casn’t document the black box that is the neural net that solvbes the problemn.
“If this is true—if evolution can produce something as simple as a sorting program which is fundamentally incomprehensible—it does not bode well for our prospects of ever understanding the human brain.
‘twould be cool if this was automagically filled in from the link in the ereader…. sigh…
I have used mathematical tests to prove that the evolved sorting programs are flawless sorters, but I have even more faith in the process that produced them than in the mathematical tests. This is because I know that each of the evolved sorting programs descends from a long line of programs whose survival depended on being able to sort.
The fact that evolved software cannot always be understood makes some people nervous about using it in real applications, but I think this nervousness is founded on false assumptions.”
My thoughts on why I linked this goes here…:
“The amazing thing to me is not that a computer can hold the contents of all the books in a library but that it can notice relationships between the concepts described in the books—not that it can display a picture of a bird in flight or a galaxy spinning but that it can imagine and predict the consequences of the physical laws that create these wonders. The computer is not just an advanced calculator or camera or paintbrush; rather, it is a device that accelerates and extends our processes of thought. It is an imagination machine, which starts with the ideas we put into it and takes them farther than we ever could have taken them on our own”
My thoughts on why I linked this goes here…:
“Mathematicians had been thinking about computation and logic for centuries but—in one of the more dazzling examples of synchrony in science—Turing, Church, and another British mathematician named Emil Post all independently invented the idea of universal computation at roughly the same time. They had very different ways of describing it, but they all published their results in 1937, setting the stage for the computer revolution soon to follow.)”
My thoughts on why I linked this goes here…:
“A second reason not to expect a perfect computer is that most computer failures are not caused by incorrect operation of the logic. They stem from errors in design—usually in the design of the software. Programmed computers, including their software, are by far the most complex systems ever designed by human beings. The number of interacting components in a computer is orders of magnitude larger than the number of components in the most complex airplane. Modern engineering methods are not really up to designing objects of such complexity. A modern computer can have literally millions of logical operations going on simultaneously, and it is impossible to anticipate the consequences of every possible combination of events. The methods of functional abstraction described in the previous chapters help keep the interactions under control, but these abstractions depend on everything interacting as expected. When unanticipated interactions occur (and they do), the assumption on which the abstraction rests breaks down, and the consequences can be catastrophic. In a practical sense, the behavior of a large computer system, even if no failures occur, is sometimes unpredictable—and this is the overarching reason that it is impossible to design a perfectly reliable computer.”
Brain has massive parallelism! And aren’t there also some preprocessing of things like edges, motions, etc. that might help the brain at this? And is this what can be done with MP supercomputers? ?maybe this is a stupid -obvious statement of common knowledge“
“The computers were slow because they were sequential; they could do only one thing at a time. A computer must look at a picture pixel by pixel; by contrast, a brain perceives an entire picture instantly and can simultaneously match what it sees to every image it knows. For this reason, a human being is much faster than a computer at recognizing objects, even though the neurons in the human visual system are much slower than the transistors in the computer.”
Template below:
My thoughts on why I linked this goes here…:
“And here is the quote…”
