Hello Internet strangers (and the dozen or so friends and family I’ve guilted into reading my blog),
After months of boring everyone I know talking about wanting to start a blog, I have finally launched: AI Insanity. And you’re probably all wondering what we’re all doing here, and why I chose that name. Well, this post will attempt to answer that question, and many others.
Q: Why are you writing this blog?
A: I am writing this blog because the field of Artificial Intelligence, and the way it’s being covered by journalists, public figures, and occasionally even researchers who should know better is … Insane! I claim to be no kind of a super-expert in AI—I am a 3rd-year PhD student in Machine Learning at Carnegie Mellon working in computer vision—it is obvious that a lot of people talking very loudly about AI have no idea what they’re talking about. Well known public figures **cough** Elon Musk **cough** fear-monger about the dangers of AI, even as we continue to struggle with some of the fundamental challenges in the AI disciplines such as Computer Vision, Natural Language Processing, Machine Learning and Robotics.
For a colorful illustration, see A Compilation of Robots Falling Down at the DARPA Robotics Challenge.
To give you some context, this was a major robotics competition hosted by DARPA, perhaps the most important US funding agency for robotics research and attended by the top robotics researchers in the world including from CMU, MIT and KAIST.
By the way, I don’t think the reason that these misconceptions are so widespread is because people are not smart, but because much of what people hear about AI is from journalists who do not know anything about AI and do not learn, some very smart public figures who are nevertheless very ignorant about the field of AI, and marketing materials made by companies to sell their product, not to properly explain AI concepts. As the old saying goes: those who know do not speak and those who speak do not know.
This blog, therefore, seeks to correct this issue. When there is hype about a AI technology, I will be there to pour cold water on everyone’s heads and talk about the limitations of the work. When someone says something nuts about AI taking over the world, I will be there to set your mind at ease. No BS will escape my sight.
Q: What qualifies you to be an AI media critic exactly?
A: Harse, hypothetical questioner. As I humble bragged earlier, I am a PhD student at Carnegie Mellon University in the Machine Learning Department. I primarily work in Computer Vision, but I’ve dabbled in Natural Language, Reinforcement Learning, and Robotics and I have taken courses in a variety of topics in Machine Learning and Statistics
I also have some experience on the news side of things. At Georgia Tech, I was a news reporter, a copy editor, News Editor and Managing Editor of the Georgia Tech newspaper The Technique. I covered a wide variety of stories from hard news to university issues. I covered President Obama’s speech at Georgia Tech in 2015. Here is me with my White House press pass.
And I got hate mail from a 40-year-old lacrosse player about a story I wrote about a Tech lacrosse coach getting fired for faking invoices. I believe that getting hate from a Lacrosse guy is the ultimate bona fide for any journalist.
Q: Are you really still bragging about the White House press pass?
A: … yes …
Q: So why do you bring this all up?
A: I believe that I am in sort of a unique position as someone who not only works in an AI subfield at a major university, but also has a background in journalism. I know both what it is like to do basic AI research and what it is like to cover a story on a topic you do not know much about. Hopefully this experience leads to some interesting insights when I cover these issues in this blog.
Q: What kinds of things can I expect to see on this blog.
A: As I say, the main purpose of this blog is to talk about the misconceptions people have about the field of AI as them come up on twitter, Facebook, Instagram, mainstream media, garbage-tributary media, carrier pigeons, and engraved on stone obelisks. Much of the blog will be linking to what I believe is a bad (or maybe good) article about AI and use it as a way of explaining a fundamental concept in AI or clearing up a common misconception. I also am thinking about a number of running features such as one where I interview someone working in an AI field and talking about their research. We will see as we go what people respond to.
Q: Who should read this blog?
Obviously I want everyone to read it, but target audience is twofold. Primarily, this blog is aimed at a general audience who hears a lot about AI, but maybe doesn’t know as much they would like and are interested in the kinds of mistakes the media and prominent people make when talking about it. I am not assuming any kind of background in Machine Learning, Statistics or AI, but I also assume that people are basically smart.
I have also heard a lot of interest from my friends in the ML and CV communities about this kind of blog. Most of them are probably as frustrated as I am. I hope this will also be a fun blog for them as well, and as I said before, if your work is misrepresented in the media, I would love to hear from you.
Q: Haven’t people already done this? Why are you writing a whole blog about this?
A: Yes, as in both research and writing, basically no idea is new under the sun. I would be remiss if I didn’t talk about the good work others have done to educate the public about AI and correct misconceptions that many have about the field.
One of my favorite pieces along these lines is The Seven Deadly Sins of AI Predictions by Rodney Brooks in MIT Technology review. It is a wonderfully written article that does a great job of explaining why there has been so many misconceptions. This article is one of my inspirations for starting this blog, and don’t be surprised if I bring up a lot of the same ideas in my posts.
Amitai Etzioni and Oren Etzioni from the University of Washington have spoken frequently about the hype surrounding the field recently. They talk about this stuff a lot, but I will point to the very good article they published in Issues in Science and Technology about the things that policy makers and the public should be concerned with as the field develops. Spoilers: it’s not about computers becoming self-aware and rebelling.
Q: How can I contact you?
A: I really want this blog to be a conversation with readers, not just me yelling into the void. If you really like (or more likely really hate) a post, or you have thoughts about it, the best thing you can do is leave a comment. Depending on my schedule, I will try to engage in the comments section and answer questions and comments. Feedback is much appreciated, especially early on as I try to work out all the kinks.
Another thing I really want to do with this blog is engage with other researchers and learn about what they are doing and how their work is misrepresented. If you are such a person and have an idea for a post or want to talk about how your work is covered, you can contact me at firstname.lastname@example.org. You should also use this email if there is something else you want to ask me that is not something you can really put in a comment. I can’t promise to answer every email, but I will check it regularly.
For the most up to date contact info, see the Contact page.
Q: Hasn’t this post gone on a little long at this point.
A: Yes hypothetical interviewer. So I will leave it at that. I hope you all enjoy reading this blog as much as I will enjoy writing it.