• Welcome to the new Internet Infidels Discussion Board, formerly Talk Freethought.

Thinking about algorithms

Have you considered that there are algorithms thinking about you?
What do you have in mind?

You must not br following the news. AI is fully developed in both commercial and military analysis of people, down to individuals.

Obama pioneered it in politics tailoring political emails for individuals.

The next step will be artificial consciousness, an analog to human thinking.
 
Have you considered that there are algorithms thinking about you?
What do you have in mind?

You must not br following the news. AI is fully developed in both commercial and military analysis of people, down to individuals.

Obama pioneered it in politics tailoring political emails for individuals.

The next step will be artificial consciousness, an analog to human thinking.

That's been the next step ever since the mid-80s.

Just like manned interplanetary missions have persistently been only 15 years away since the 1960s.
 
You must not br following the news. AI is fully developed in both commercial and military analysis of people, down to individuals.

Obama pioneered it in politics tailoring political emails for individuals.

The next step will be artificial consciousness, an analog to human thinking.

That's been the next step ever since the mid-80s.

Just like manned interplanetary missions have persistently been only 15 years away since the 1960s.

There are already commercial uses for artificial neural nets. The nets can be taught functions as opposed to being programmed with an algorithm. Add adaptive features and it starts to emerge.
 
You must not br following the news. AI is fully developed in both commercial and military analysis of people, down to individuals.

Obama pioneered it in politics tailoring political emails for individuals.

The next step will be artificial consciousness, an analog to human thinking.

That's been the next step ever since the mid-80s.

Just like manned interplanetary missions have persistently been only 15 years away since the 1960s.

There are already commercial uses for artificial neural nets. The nets can be taught functions as opposed to being programmed with an algorithm. Add adaptive features and it starts to emerge.

It has been starting to emerge since the 1980s. How much longer is it going to be starting to emerge for, before it actually emerges?

It's like commercial nuclear fusion power. It's almost certain to arrive any minute now - just as it has been for the last fifty years.
 
Actually in the 80s it was AI. It never lived up to the initial hype it ended up being applying a lot of rules fast, but today as in the news AI is coming into its own 25 years later.
 
Actually in the 80s it was AI. It never lived up to the initial hype it ended up being applying a lot of rules fast, but today as in the news AI is coming into its own 25 years later.

Today the news is hype.

AI isn't living up to the hype today. Nothing has dramatic has changed - CPU cycles, Memory and Storage are all FAR cheaper today, and huge amounts of all three are being hurled at problems, but there is no really big breakthrough from an algorithmic or architectural perspective.

Stuff like IBM's Watson have some potential, but they are basically just search engines. Watson works well with a highly structured database of accurate facts to draw on, as it showed when it won a special edition of Jeopardy. But working with a poorly structured information base that contains facts, conjectures, half-truths, random noise, falsehoods, errors and outright lies (such as what we fondly know as 'reality') needs a human to be effective - and will for the foreseeable future. The kinds of breakthroughs needed to make AI genuinely useful have simply not been made.

What we are being promised:
[YOUTUBE]https://www.youtube.com/watch?v=WFR3lOm_xhE[/YOUTUBE]

What we actually get:
[YOUTUBE]https://www.youtube.com/watch?v=NMS2VnDveP8[/YOUTUBE]
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.

What you are calling 'AI' is clearly not what most people in IT today mean by AI.
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.

Glad you recognise you were using hyperbole, if not outright misleading wording, when you stated, a few posts up, that "[t]he next step will be artificial consciousness, an analog to human thinking".
 
Laughing at the video. I've gotten that way with telephone voice recognition systems. You'd think they'd teach them swears so they'd know when to link you to a human!
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.

Glad you recognise you were using hyperbole, if not outright misleading wording, when you stated, a few posts up, that "[t]he next step will be artificial consciousness, an analog to human thinking".

The next step logically would be artificiial consiousness, an analog to the human mimd/brain. No hyperbole. Goedel speculated that it may not be possible to design one as a rule based function, but it may be possible to grow a neural net much as a human grows. I expect AC research is well under way.
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.

What you are calling 'AI' is clearly not what most people in IT today mean by AI.

AI articial intelleigence. Saying agian, back in the 80s the idea was to reduce accumulated techical experince and reduce it to an algorithm. My main design tool was Cadence for circuit design and pcb design. Printed circuit board design across many layers was initially more an art leaned by experience. The Cadence/Aleggo toll has hundreds of checks according to what we call best design practces.

You have to separate scifi from scoence.I have my own license for a 3D mechanical design tool called Alibre. The embedded AI engine allows me to do a lot of mechanical design that used to require a lot of experince. It tells me if there are problems with the design.

That is AI from an enginering perspective. Artificial consiousness to me would be asking an android to design a machine and have it creatively do the work as a human would. ST Data is an example.
 
Ok here's my program that takes a word or phrase and translates it into a phone number, complete with all algorithms. It is a command line program, you would type "phonltr" + the phrase to translate into numbers.

$COMPILE EXE
DEFINT a-z
clp$ = COMMAND$
IF RTRIM$(clp$) = "" THEN
CLS
PRINT
PRINT "USAGE:"
PRINT "PHONLTR + The letters or words comprising the phone number."
PRINT "EXAMPLE: PHONLTR deliver"
PRINT
END
END IF

strlen% = LEN(clp$)
IF LEFT$(clp$,1) = "1" THEN clp$ = RIGHT$(clp$,strlen%-1)

FOR j% = 1 TO LEN(clp$)
letter$ = MID$(clp$,j%,1)
get_number letter$
NEXT

result$ = RTRIM$(result$)

SELECT CASE LEN(result$)
CASE 7
result$ = LEFT$(result$,3)+"-"+RIGHT$(result$,4)
CASE 10
result$ ="1-"+ LEFT$(result$,3)+"-"+MID$(result$,4,3)+"-"+RIGHT$(result$,4)
CASE ELSE
FOR j% = 1 TO LEN(result$)
t$ = MID$(result$,j%,1)
IF j% <> LEN(result$) THEN
target$ = target$+t$+"-"
ELSE
target$ = target$+t$
END IF
NEXT
result$ = target$
PRINT "The phone letters used generate a non-standard phone number length."
END SELECT

PRINT
PRINT "The number is ";result$
PRINT "--------------"+STRING$(LEN(result$),"-")

END

SUB get_number(letter$)

SHARED result$
letter$ = UCASE$(letter$)

SELECT CASE letter$
CASE "-"
EXIT SUB
CASE "1","2","3","4","5","6","7","8","9","0"
number$ = letter$
CASE "A","B","C"
number$ = "2"
CASE "D","E","F"
number$ = "3"
CASE "G","H","I"
number$ = "4"
CASE "J","K","L"
number$ = "5"
CASE "M","N","O"
number$ = "6"
CASE "P","Q","R","S"
number$ = "7"
CASE "T","U","V"
number$ = "8"
CASE "W","X","Y","Z"
number$ = "9"
CASE ELSE
number$ = number$
END SELECT
result$ = result$ + number$
EXIT SUB

END SUB
 
Au contraire amigo.

AI is well developed in engineering computer aided design. Especially in mechanical design, electronic design, and printed circuit board design. A lot of design work was not possible until AI coupled with fast cheap computers arrived. AI never met the i9ntial hype, but it is an integral part of technology development, like small cell phones.

It does not replace human thought and invention, but it applies accumulated knowledge and empirical experience in ways beyond human capacity. Back in the 80s I would have thought by now electrical engineering would have been replaced by AI, if AI met the hype.

Glad you recognise you were using hyperbole, if not outright misleading wording, when you stated, a few posts up, that "[t]he next step will be artificial consciousness, an analog to human thinking".

The next step logically would be artificiial consiousness, an analog to the human mimd/brain. No hyperbole. Goedel speculated that it may not be possible to design one as a rule based function, but it may be possible to grow a neural net much as a human grows. I expect AC research is well under way.

In the meantime, people who actually work in AI hardcode long lists of exceptions to be treated differently, because those self-learning systems still don't always get basic relations between entities after reading the entire Wikipedia in five different languages, and then some, for context.
 
Back
Top Bottom