Spot the Algorithm
We're all familiar with the stereotypical sci-fi horror scenarios in which robots take over the world. They can be entertaining, intellectually stimulating, and even believably portending (see Black Mirror's Metalhead). Good sci-fi makes us think, in spite of how well or poorly it predicts the details of the future.
This post, however, isn't so much about robots-as-hardware, but as software. Perhaps because of language, we don't usually think of algorithms as robots. Perhaps software shunts our vision to a more abstract conception than hardware. Yes, the term 'bots' has gained awareness in mainstream consciousness. But the idea of software robots doesn't strike us with quite the vivid conception that The Terminator does.
Given the ever-infiltrating predominance of algorithms in our businesses, economies, social and governmental infrastructures, and personal lives, I think we need to better arm ourselves in understanding that the 'future' is now when it comes to software robots - more precisely the programs that execute mathematical algorithms. There are ethical, moral, and survival dimensions that we can't afford to ignore.
So let's apply Isaac Asimov's Three Laws of Robotics and apply them to algorithms, switching-out "robot" with "algorithm" and see what things look like (example consideration in brackets):
- An algorithm may not injure a human being or, through inaction, allow a human being to come to harm. [Insurance company policy denies claim of pediatric cancer that can only be treated by one specific orphan drug.]
- An algorithm must obey orders given it by human beings except where such orders would conflict with the First Law. [Algorithm, by this rule, should override policy.]
- An algorithm must protect its own existence as long as such protection does not conflict with the First or Second Law. [Algorithm executes code to protect itself from the insurance company's developers' override of the override]
Granted, there's some stretch in the examples and Asimov's rules aren't the most robust algorithm to determine how to build 'safe' robots, but the point here is to get a better view of what we mean when we talk about 'bots'. Soon, they'll be everywhere, controlling entire financial markets, making diagnostic and treatment decisions, driving cars, playing cherubim, etc. In a sense, they already are doing those things, but they'll continue to get "better, stronger, faster" - that is more 'technologic' (for you Daft Punk fans).
It's not that we're facing doom and gloom. Rather, we are facing crucial problems. We not only have questions to answer, but we also have questions we need to have the imagination to ask. The doom and gloom outcome depends on what we do or not do now.
The words we choose to help us make sense of the world - a rapidly changing and disorienting one at that - are absolutely relevant to how we think. We are all-to-deceived and distracted from the present dangers and opportunities that the words "robot" and "bot" pose. If we truly value the role Technology and want to direct its drive more than the other way around, we should not let misplaced language divert our eyes. We don't need to be alarmists, but we need to be aware.
Mind the robot, but...
Spot the algorithm.
Database Associate at a Cincinnati Nonprofit
7 年There are a number of algorithms that relate to the job search. Only one of them is a computer program. All of them have yielded less than satisfying results as it relates to my not yet having a career. The computer algorithm is of course the ATS, which you don't have to go any farther than LinkedIn to know about the shortcomings of that software. It's really draining to send out customized, keyword-laden resumes, and still not get interviews for positions you almost perfectly match. The other job search algorithms I'm aware of are not computer programs. One is algorithm of going to college, selecting a major, applying to jobs within that major the semester before you graduate, and right after you graduate, interviewing during your last semester and the weeks following graduation, and having one or more job offers after graduating. I went to grad school determined to get it right following not being able to get a job with a bachelor's degree, and I did all of these things, including getting a great internship, and I'm still getting the same job search results as I did before going to grad school, while watching my classmates thrive, in many cases having to do much less to get their jobs than I've been doing in my job search.
Telecommunications Professional
7 年we dont need software robot at all, those robot are man made machine which amplify both human mind and strength and mind part including both advance thinking and weakness part, even they do amplify dark side more, so the software robot is machine human which be doom..and we need ai which is simply advance thinking tools for great help, they are completed different from human who does not greed/cheating/ignorant/se7en
GSA AI Community of Practice, NLP lead. Treasurer, Asian Financial Society CFA AMP
7 年It may be instructive to note that the Three Laws of Robotics are themselves a type of meta-algorithm. In other words, the only effective constraints on algorithms are other algorithms, because human reaction time is too slow. Look at the last 'flash crash.' Exchanges have 'circuit breakers' for this reason in part. As we turn over more of our daily lives to algorithms, it behooves us to build in safeguards using independent software agents.