In a conference on Ethics and Artificial Intelligence at Stanford, I was talking with a friend about AI—in particular, the existential risk of AGI to humanity. The whole “Bostrom” mentality, for me, was a bit of a struggle…not because I didn’t think it possible, but rather that the possibility of AI threatening humanity seems very remote. As I spoke to my peers, we found ourselves discussing the question, “What is the real danger?” While I still would argue that AGI is less of a danger than some might imagine, I quickly came to the conclusion that we face an existential risk now, and it is ever-present, likely and extremely dystopian.
I ended up writing and submitting a paper for peer review on this exact topic. The paper began as a writing exercise, but I ended up actually making a working version of this machine horror with off-the-shelf technology available now.
Imagine this scenario: The government creates a national ID, but this ID is a cryptographically secure bracelet of stainless steel. It does everything from being a credit card to truly “helping” you with your right thinking. With a president like Trump helping you he can now encourage or at least ‘negatively’ reinforce poor decisions, such as shopping at the wrong stores or voting for the wrong person. Imagine the government helping with all sorts of decisions and negatively reinforcing the bad choices, with the generous application of 250 volts applied to help you achieve that right thinking. On the positive side, version 2 also comes with a Bluetooth subdermal implant that gives you a nice hit of MDMA (the drug commonly called Ecstasy) to reinforce those positive choices.
The best part was this is available now with ready-made and available technology. It took all of ten minutes to hack such a shock bracelet—and a friend at the University of Utah helped with the details of the subdermal implant.
So, there you go: six weeks with off-the-shelf technology and we had a working version of this super awesome tool ready for Trump or any other leader that can help you with your right thinking. What a great plan, you might say…but wait, there’s more: With-off-the shelf technology like Machine Learning, we can help lift people out of their ignorance with the manipulation of elections and take things a step further too (for the greater good, of course). Maybe some thought police action? Or something else fun like that? This is what can be done now, and it being combined with humanity and all our weaknesses is the real existential risk. In my opinion, this abuse needs to be fought at every turn unless you want a Responsibility Bracelet.
If you’re interested in the details of our working Responsibility Bracelet prototype read the paper Feasibility Study and Practical Applications Using Independent Core Observer Model AGI Systems for Behavioral Modification in Recalcitrant Populations; found in the book found at https://www.springer.com/us/book/9783319993157.
Remember, Big Brother is watching…and we can only escape if we all work together.