 So the UNM Engineering Student Success Center asked me to present a brief history of computer science Because I used to be a UNM computer science professor and because I'm old and today I'm gonna focus on the old stuff when we get to PCs and smartphones. I'll leave it up to you Now they told me I should be light-hearted and fun like a one-man John Oliver show Wait The history of computer science the history of computer science for the most part is the history of digital computing Which begins with the breakthrough discovery of the digits? They're great for counting things and they come in fives and tens which is just the way we like to count Incredible luck or alien influence you decide in any case counting is good for lots of things Like to count up how much you owe me says the king the high priest the local big shot and don't you forget it? And that was actually a problem with early digital computing fingers were convenient but you couldn't count very high unless you had lots of friends and The counts kept getting erased when you decided to use your fingers for something else But information technology research and development was on the job Counting on fingers led to using pebbles or sticks or stones or shells to keep the count and Coming up with all sorts of tally marks like these modern ones. Look at those cute little hands one two three four five one two three four five and Check out these marks that were scratched into the leg bone of a baboon 20,000 years ago. That's persistent memory now granted this may not be the baboon's favorite memory technology But there were more innovations to be had like clay It's easy to mark on wet clay then let it dry if you wanted to you could wet it again and smooth it over like a Sumerian etch a sketch But if you cooked it in a hot oven the clay tablet turned to frickin stone and the royals and accountants everywhere like that a lot and The resulting persistent memory market was basically owned by big clay for centuries And clay got used for everything like this from 5,000 years ago. What was important enough to bake this tablet? Beer who gets the beer? Well, I don't care if your uncle is Sargon of a cod It says right here. You only get one vessel of beer. Good day, sir Or this little note it's three three or four thousand years old It's a customer complaint letter and you thought it was tough dealing with Wells Fargo Yeah, I have a friend a new sender So counting up is great and right once memory is great But for more complex math, we really want persistent but rewritable memory and Progress marches on with this sturdy wooden counting frame You can write and rewrite numbers all day long or check out this streamlined monster It's got like 17 hands one two three four five One two three four five six one two. Whoa, how's it work? One two three four five six seven eight nine Counted to nine with one hand. I use these fingers twice can't do that with right once memory So with a device like this if you knew the bead the twiddling rules and the algorithm You could do complicated things like multiplication, which is good for figuring out how much Everybody owes me all at once So put the memory device together with the instructions and you get a digital computer, right? Well, not quite you have to plug them into a human or they don't work Memory by itself is just all passive-degressive. I'm just repeating what you told me and the rules Just give orders you do this you do that the rules never do anything themselves So we need something else to examine the situation make a decision and then turn the crank and move on It seemed obvious that there were only two choices teach a human the rules and have them do whatever it is or Design the rules explicitly into the machine up front so it can't not do them but then the rules can't be changed later and That was the great split in digital computing You could build machines and then try to make them flexible or you could deploy on humans and then try to keep control opportunities and risks either way Now when humans are writing rules for other humans things can get really complicated really fast And in some ways that's the history of science people programming people is also culture and parenting and law But people programming people to do math physics science has proved to be especially powerful Royalty likes that Now the math humans did kind of get into sort of arms races with each other about notation Trying to pack the most power into the shortest program and utterly baffling everybody else in the process I know that one No Who knows Now there were also many efforts over the centuries to make automatic Deciders of increasing flexibility and speed here's the pascaline of 1652 an adding machine designed by blaze Pascal And people made automata machines that moved on their own and sort of acted like alive They weren't really ever useful, but then a lot of them were made to please the royalty This is a view of Vauconson's digesting duck from 1739 which appeared to eat grain drink water and poop The mechanical Turk claimed to be a chess playing machine Built to impress Maria Theresa the Empress of Austria It played chess really well, but it was a fake with a human chess player hidden inside as the decider It toured for decades anyway In Japan there was a long tradition of caracuri puppet automata that did things like move and bow to serve tea Yes, rich people bought them On the other hand the Jacquard loom of 1804 which could weave intricate patterns from punch card programs was a success indeed by 1812 there were eleven thousand of them at work in France and By the way, the Jacquard machine design built on previous loom work by among others Vauconson the digesting duck guy incredible luck or alien influence you decide Here's a look at the programming cards and here's the man himself not in a lithograph or an etching But woven in silk using 24,000 punch cards on a Jacquard loom Charles Babbage sometimes called the father of the computer owned one of these silks and was Reputedly inspired by it. So over the centuries there was progress on the quest for flexibility and fast mechanized deciders and Development of the programming languages they accepted we even started to get software documentation and All the digital computing concepts had been in the air one way or another since the beginning as Much as anything it was a matter of multiple technologies Matureing plus the will the investment and the audacity to make it all work Several machines can claim to be the first electronic computer depending on what's considered essential in Bletchley Park north of London the Colossus mark 2 was built specifically for code breaking The ENIAC funded by the U.S. Army and built at the University of Pennsylvania was more flexible than Colossus But programming it involved physically recabeling and rebuilding parts of it and took a long time But after that it could do a thousand additions five thousand additions every second And over the next couple of decades electronic computing exploded around the world the U.S. Army funded several more machines With pieces of them shown here from the left a board from the ENIAC the Ed vac the Ord vac and the burl ask one on the right And just to be clear the folks holding the boards are mathematicians programmers project staff bad-ass nerds in heels and It was an age a time of fantastic progress for the math humans too There was finally a non-human machine with great rule following flexibility and reliability Thousands of algorithms and variations were explored just to pick one out of that hat the Patricia tree algorithm was created by Don Morrison who was a member of the UNM computer science department way back in 1968 Commercial computing was on the rise as well Big machines epitomized by the IBM system 360 which debuted in 1965 And this is a deck system 10 my favorite computer of all time not this exact one But I've loved a few The deck 10 was one of the first popular Time-sharing machine so that a lot of people could use a piece of a machine They'd never be able to afford all on their own I loved learning how to use it and I'm sorry, but Tico is still the most awesome text editor ever fight me Program, I love programming it too, and I did it a lot And it this great job. I had as an undergrad was being a system operator for a deck 10 I Didn't dwell on it at the time, but you know this was a huge magnificent machine Like most mainframes it had a private room all its own really it wasn't in the room so much as it was the room the machine room thick doors always locked raised white panel floors with cables running in every direction underneath Smooth even lighting AC humming always cool The printer chattering as the next job came out the blinking lights on the main console right out of Star Trek Usually not that much to do unless there was some problem there then cabinet after blue cabinet of main memory Each one holding I think about a megabyte Although deck the deck 10 didn't count bytes quite the way we do now and then further back the tape drives That would meet you face-to-face as you change the reels for one of the users or another And there were always users users users day or night pack cheek to jowl inside the memory of this incredible machine Doing homework playing games analyzing physics experiments Or just messing around to see with the machine to see what it could do I did some of that, you know Colossus didn't have to worry about randos trying to provoke it into doing something weird neither did eniac and Getting everything perfectly safe turned out to be kind of hard. We haven't figured it out yet And the locks and the thick doors kind of missed the point It's really about the cables from all over campus Converging under the white panel raised floor and punching straight up into the brainstem of the machine So its user names and passwords special operator privileges Hierarchy was imposed but no matter The machine utterly trusted its program. That's the source of its flexibility and it could not do otherwise the brief age of Secure computing was over The deck tens lasted into the 80s, which was a good run for at the rate machines were evolving One was in fact decommissioned while I was in grad school. This was one of my my treasured possessions It wasn't security problems so much as just the company wanted to concentrate on its newer VAC systems But things were changing in lots of ways anyway So That's it. I didn't want to end on a downer note But the quest for the fully automated totally flexible mechanical decider Really has succeeded too well We need to focus on robustness So I published this argument in in 2013 the future of computer science is robust first and you know, of course the consensus now on the lessons of 2026 is that you know, we pretty solidly missed robust first But you know robust later is better than robust never wait What year is this? Thank you any questions