Igor KulmanYears ago I wrote a Pascal interpreter in #swiftlang so I thought, why not try making it an iOS app. Pascal is still taught at some school as the first programming language and the retro theme is just so nostalgic. #iosdev
Archimage@igorkulman @krzyzanowskim Neat. I wish Apple still supported non-Apple languages out-of-the box. Would prefer Pascal/Modula-2 over Swift. ;)
Marcin Krzyzanowski@krzyzanowskim @igorkulman lol. Is this a competition to see who can name the most obscure language theyβve actually used? Yup, played with Ada. ;)
Alan Langford π¨π¦π§€π§ζ@SergKoren @krzyzanowskim @igorkulman Old languages: Algol, FORTRAN IV and V, APL, Pl/1, COBOL.
Helped getting a Pascal compiler running on our mainframe sometime around 1976, was on the standards committee up here for Ada, but not for long... it was a rubber stamp operation and boring as hell.
Helped getting a Pascal compiler running on our mainframe sometime around 1976, was on the standards committee up here for Ada, but not for long... it was a rubber stamp operation and boring as hell.
@SergKoren @krzyzanowskim @igorkulman Hey, at the time I was working on early AI: a recursive descent natural language parser written in Pascal, and the first work on neural nets had already been published by then.
It's more accurate to say before a bunch of VC's started pushing AI as a product when it should still be in research labs until it works well enough to not be a force of destruction.
@alan @krzyzanowskim @igorkulman Neat. I wrote on of the first chatbots on Compuserve, which pretended to be a user in chat. I guess that would be considered AI nowadays :/
@SergKoren @krzyzanowskim @igorkulman All depends on whether or not there was some sort of inference engine in there. Otherwise it's just a really cool FSM.
@alan @krzyzanowskim @igorkulman yeah, it was a combination of FSM with rudimentary language brute-force processing. Not very "intelligent", but could pretend to converse and maintain a bit of conversational context. Also, had a bit of randomness thrown in to initiate chats.
@SergKoren @krzyzanowskim @igorkulman It's funny. In "The Emperor's New Mind", Roger Penrose attempted to prove that AGI was computationally impossible, and my response to his hypothesis was "nothing a random number generator couldn't fix."
@alan @krzyzanowskim @igorkulman lol. What matters is that human behavior appears random to other humans (even if it isn't computationally so).