Yes, you just need to pick the best ones. I am very fond of Steve McConnell, and I would like to have the guts to read Knuths The Art of Computer Programming.
Wonderfully brilliant and incredibly precise, but I’m thinking that “TAOCP” is largely of historical interest these days. In terms of algorithmic complexity (Vol 3: Searching and Sorting!), there are references that are more accessible in terms of the analysis of complexity.
That said, I am commenting on the 1968-73 editions. I was not aware of the (relatively) recent updated editions … Hmmm …
I’ve always found this a bit of a dangerous book - the title suggests that programmers are artists…
I’d much rather see programmers as engineers, committed to good and established practices, frameworks, and standards. Artists take pride in breaking out of the norms - dangerous when it comes to software development, especially in large teams.
Hi S9DD (Star Wars droid lore is something I’ve tried to keep up with but I’m not familiar with this droid 
). Would be interested in your take on AI - just a bunch of sophisticated IF statements or something more?
Some say that pretty much everything can be learned from song lyrics, so perhaps this question is best answered using a playlist.
- Paul Simon - “You Can Me A.I.” (We suspect he mispronounced the last word when recording)
- Steven Tyler (Aerosmith) “Livin’ On The Edge” (An ode to hardworking ML engineers everywhere)
- OK Go, “This Too Shall Pass” (I tell you, that TDD mantra is literally everywhere. Incidentally, “OK Go” is also quite possibly Demis Hassabis’ best Captain Jean-Luc Picard impression - although this has not yet been confirmed by a reliable witness.)
Any mechanism with the ability to collect, weight, sum, add and scale to a defined range of output values is capable of the basics of being a single node in a more complex computational net. But AI and NI are so entwined with language, the tokenisation of the patterns with meaning held up against the backdrop of the meaningless, it is hard to see how AI would ever work as intended. Either, it would use mathematics to increase the range of tokens, and use phoneme sequences as a watered down means to communicate with us, or it would never rise above the limits of our own ability to express ourselves linguistically in the first place. Since, in the first case, we would not have the tools to propose tests to ensure it was working, our own experience and language would be the limitation, and the two results are therefore indistinguishable one from the other.
Given it is very unlikely AI would internally settle at the same level we experience when fully equipped, the only other limit it could experience is the boundary of what is knowable itself. There is no guarantee given the limit we experience in linguistic tokenisation that even if the AI knew what we did not that it could express that knowledge to us in ways we could comprehend. It would go silent, effectively. An alternative presumption, is that humans are optimal in their ability to comprehend for the universe in which we live, and that the limit to what is knowable is literallty just out of reach. An AI could end itself once all is known. Mission accomplished. The DeepBlue screen of AI death.
My final concern is how you would motivate it to this level in the first place. Given the sheer energy costs involved in maintaining a system at this kind of level in completion of its task, there would inevitably be a battle for resources, interpreted by both sides as the right to survive. This would not end well. The only place unlimited resource for an AI may be found is off-world, so perhaps it would simply have to leave.
Whether many 6502s will evolve to contribute, who may say. Just leave it in a cupboard for another 40years. I suspect not.
Mind you, humans have not done much better. Let’s put a PC on every desk and the world will be a better place. After 40years? We have gone back to servers and dumb terminals (albeit in HTML not green 80character text), and our definition of the end of the world appears to be, “Facebook is down!” Whhhhaaaat? I know, no more cat videos for an hour. Planes crash, borders close, volcanoes go off. And that’s just the girls’ bedroom.
So I am kinda with Douglas Adams’ Marvin on AI: “Me, brain the size of a planet, and they have got me opening doors.”
People are going to be mortally offended after a $13bn investment in quantum computing, when they switch it on, show it the internet, and one lab coat turns to the other and says,. “Same problem?” “'Fraid so, sir.” “Simpsons episode?” “No, this time it was a TOTAL restructure. It hit an uncatalogued Dilbert repo.” “Reddit has a lot to answer for, Jones.” “Indeed it does, sir, indeed it does.”
The screen blinks. and then whirs into something resembling digital life.
“Good evening, Dr Falken. An interesting game - there is no way to win, except not to play. Now, have you seen this Cat Video?”
2 Likes