10.24
In Future , Technological Singularity , Videos | Tags: Google, google tech talk, human computation
WikiPedia explains human computation as,
… a computer science technique in which a computational process performs its function by outsourcing certain steps to humans (Kosorukoff, 2001). This approach leverages differences in abilities and alternative costs between humans and computer agents to achieve symbiotic human-computer interaction.
More plainly, human computation is a system that uses human decision making where questions require more or higher reasoning than any computer program can accomplish efficiently. Efficiency is the operative concept here. It’s possible to write programs of infinite complexity that make strategic decisions based on vast amounts of statistical data; but our species hasn’t completely mastered the physical hardware to do this. If-then logic is extremely powerful as applied to reason, but in the programming sense, only when supplemented by scores of data.
The human memory still holds a lot of data that isn’t available online–or that hasn’t yet been entered into a database, at least. We think of companies like Google compulsively hoarding and processing tons of data, but the human brain still holds more and processes more information than this organization could fathom. Facial recognition is a perfect example of this because, while a video camera can record you and a computer can store this data, all that is collected is the series of images making up the video. You can’t, for instance, query a JPG to know the manufacture of the shirt on the news anchor.
In 20th century video surveillance, the software didn’t yet know who the person was on camera. It still took a human to answer this question because we could do it faster. The relationships and associations human beings connect to bits of data, and the manners we employ to retrieve or “remember” these data are inherently efficient. We’re teaching the internet how to be like this, too.
Cheaper hardware, more advanced programming methodologies, and more efficient database systems are making it possible for our species to record data about physical space that hasn’t before been logged. Take for instance Facebook’s photo tagging and the similar service now offered by flikr. How far off are we from our homes knowing that it’s us at the door? How long before Facebook’s upload tool automatically recognizes everyone in a photo?
I came across the following video some time ago. The idea is actually pretty simple. Use a program to handle mundane tasks. Treat decision making in code where enough data and processing power exist. Use humans when judgment is required. Judgment, unlike logic, implies some deficiency in knowledge of a given occurrence–some degree of faith, character or experience; whereas logic suggests a definite outcome for a definite circumstance. Humans do this far better than any combination of hardware and software. Check out the video below. You might learn something.
0 ResponsesLeave a comment ?