Artificial Intelligence: Once Again Safe to Use the Term
July 11, 2008
For years, I have been reluctant to use the phrase “artificial intelligence”. I danced around the subject with “computational intelligence,” “smart software,” and “machine intelligence.” Google aced me with its use of the term “janitors” to refer to a smart factory that generated smart digital robots to clean up messes left from other data processes.
Now, the highly-regarded Silicon.com has made it okay for me to say, “artificial intelligence” and use the acronym “AI” without fear of backlash. Tim Ferguson has an essay “Artificial Intelligence–Alive and Kicking” built in part upon an interview with Professor Nigel Shadbolt, head of artificial intelligence at Southampton University.
The point of the essay is that AI continues to thrive without the wackiness that accompanied the hype from a decade ago. Examples of practical AI are voice recognition (my phone often doesn’t understand me when I am driving) and vision processing software (works okay for certain types of object recognition but not others).
The key point in the essay was this statement attributed to Profession Shadbolt:
What we’re seeing with the web is the way in which it can bring those things that computers are good at in co-ordination with what people are good at. You use people’s innate intelligence and ability, you connect them up on planetary scale and you’ve got a new kind of assisted intelligence. It isn’t an AI because it’s not in any way self aware – but it’s a phenomenal, powerful thing.
Bingo. What people are calling social search, collaboration, and intelligent systems is a mash up. I quite like the phrase “assisted intelligence.” Software can be more intelligent when the inputs and outputs of humans are factored into the probabilities used by algorithms to “decide”.
I will promptly co- opt the phrase “assisted intelligence”. I will give Messrs Ferguson and Shadbolt credit in this essay. I know that subsequent uses will be less disciplined about giving credit where credit is due. “Assisted intelligence” is a useful coinage.
I would like to offer three observations, which is my prerogative in my own personal Web log:
- Artificial or assisted intelligence is going to require a heck of a lot of resources, particularly if the volume of digital information continues to go up. How many companies will have the appetite to craft a large-scale system. Certainly police and intelligence authorities, companies like Google and Microsoft, and giant multi-nationals like big pharma.
- The spectrum of AI applications will range from the mundane (my thermostat adjusting itself to keep my environment a constant 72 degrees Fahrenheit to the exotic (the aforementioned janitors doing the work of human subject matter experts inside Google’s data centers). At the same time we become indifferent to AI, some applications will make headlines. There will be some debate of artificial and assisted intelligence going forward.
- AI (both assisted and artificial) will disintermediate some people along the way. Life will be good for wizards and rocket scientists. Life will not be so good for those displaced; for example, why would a start up publisher want to use the job descriptions for a traditional printed newspaper publisher. Better to trim the staff, focus on software, and keep the costs low and the margins as high as possible.
AI is back. I don’t think it ever left. The media veered into more trendy subjects. Let the applications flow.
Stephen Arnold, July 11, 2008
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
