Having worked with computers since they were first invented (I am that old!) I am amused by the current use of the terms AI and artificial intelligence. Take any recent article about AI and replace "artificial Intelligence" with "computer" and the article remains essentially the same. Transcription services have been around for decades, Dragon NaturallySpeaking, for example.
I attended voice recognition research seminars in graduate school in the early 1970s. I used Dragon to write emails to my son in college in the 2002. A friend of mine transcribed many hours of interviews with it for his Ph.D. thesis 25 years ago. A detailed but relatively long article about the history of AI is at: https://en.wikipedia.org/wiki/History_of_artificial_intelligence.
"AI" was called "expert systems" in the 1990s. They were often designed to help doctors with diagnoses and businesses with decision making. I used one and it was very convenient, but a bit cumbersome, like a fast Wikipedia but faster than an encyclopedia or medical textbook.
But people still complain about some very funny translation errors when they speak to their iPhones.
In 1962, in his book “Profiles of the Future: An Inquiry into the Limits of the Possible”, science fiction writer Arthur C. Clarke formulated his best-known and most widely cited comment: “Any sufficiently advanced technology is indistinguishable from magic”. See some perceptive quotes about AI and technology at: https://www.brainyquote.com/quotes/arthur_c_clarke_101182
Perhaps we are still at that stage of "AI" now. Are we mistaking the output of large server farms, huge database systems, small computer chip sizes, sophisticated software, and high speed electronic machines as thought?
AI is inevitable. The greatest potential problem that I see is accuracy, not just in spelling or word choice, but in facts. Accuracy in spelling and word choice can be learned. Dragon software had you reading sample paragraphs to learn your speech patterns. It also offered variations of the software to accommodate regional speech patterns, accents and colloquial terms, and it offered software for specific professional vocabulary such as medical. Pretty amazing stuff at the time, and it was a boon for transcription.
But, how does AI present facts? Those facts have to come from somewhere. Who or what provides them? AI providing facts to AI is the snake eating its tail. Today’s cartoon exemplifies that. AI in fiction is lazy, although it might improve the dreadful grammar I frequently see in today’s fiction. AI in transcription is a great timesaver. AI in journalism is the sticky point. Transcription, great. Reporting facts without human input, not so much. Joe Friday would be appalled. “Just the facts, ma’am.”
Why not credit AI when it contributes to a story the same as you would any other author? We see this line: "xxx, yyy, and zzz, contributed to this article" in many publications. Why not simply add AI as one of the names in the list. Where it gets sticky is when an author does NOT credit his or her co-authors, whether they be AI or Woodward and Berstein.
Having worked with computers since they were first invented (I am that old!) I am amused by the current use of the terms AI and artificial intelligence. Take any recent article about AI and replace "artificial Intelligence" with "computer" and the article remains essentially the same. Transcription services have been around for decades, Dragon NaturallySpeaking, for example.
I attended voice recognition research seminars in graduate school in the early 1970s. I used Dragon to write emails to my son in college in the 2002. A friend of mine transcribed many hours of interviews with it for his Ph.D. thesis 25 years ago. A detailed but relatively long article about the history of AI is at: https://en.wikipedia.org/wiki/History_of_artificial_intelligence.
"AI" was called "expert systems" in the 1990s. They were often designed to help doctors with diagnoses and businesses with decision making. I used one and it was very convenient, but a bit cumbersome, like a fast Wikipedia but faster than an encyclopedia or medical textbook.
But people still complain about some very funny translation errors when they speak to their iPhones.
In 1962, in his book “Profiles of the Future: An Inquiry into the Limits of the Possible”, science fiction writer Arthur C. Clarke formulated his best-known and most widely cited comment: “Any sufficiently advanced technology is indistinguishable from magic”. See some perceptive quotes about AI and technology at: https://www.brainyquote.com/quotes/arthur_c_clarke_101182
Perhaps we are still at that stage of "AI" now. Are we mistaking the output of large server farms, huge database systems, small computer chip sizes, sophisticated software, and high speed electronic machines as thought?
AI is inevitable. The greatest potential problem that I see is accuracy, not just in spelling or word choice, but in facts. Accuracy in spelling and word choice can be learned. Dragon software had you reading sample paragraphs to learn your speech patterns. It also offered variations of the software to accommodate regional speech patterns, accents and colloquial terms, and it offered software for specific professional vocabulary such as medical. Pretty amazing stuff at the time, and it was a boon for transcription.
But, how does AI present facts? Those facts have to come from somewhere. Who or what provides them? AI providing facts to AI is the snake eating its tail. Today’s cartoon exemplifies that. AI in fiction is lazy, although it might improve the dreadful grammar I frequently see in today’s fiction. AI in transcription is a great timesaver. AI in journalism is the sticky point. Transcription, great. Reporting facts without human input, not so much. Joe Friday would be appalled. “Just the facts, ma’am.”
Why not credit AI when it contributes to a story the same as you would any other author? We see this line: "xxx, yyy, and zzz, contributed to this article" in many publications. Why not simply add AI as one of the names in the list. Where it gets sticky is when an author does NOT credit his or her co-authors, whether they be AI or Woodward and Berstein.