Four things going on in the world of AI that keep me up at night

Photo by Franki Chamaki on Unsplash

I love lists. I love lists so much that I treat it like a personality trait. I have daily to-do lists that I keep in a notebook, as well as a smattering of numbered lists on sticky notes above my desk. I have shopping lists, bucket lists, and lists of lists taped all over the walls of my apartment.

There are probably a couple reasons behind this passion of mine. I like my information organized, and moreso, it’s visual prioritization: if something’s important, it will make the cut. If it’s not something that’s notable, I don’t want it anyway.

I have spent hundreds of hours reading and watching videos and listening to podcasts about Artificial Intelligence. So of course, I’m going to do what I do best: make a list about it.

Here are the top four things on my (constantly growing) list of important things going on in the world of AI. At least, they seem important to me.

1. Driverless cars— because my personal dream vehicle is one that involves the least amount of effort.

So I don’t want to be too predictable, but as someone who genuinely hates driving, this concept is so intriguing to me. I always get so stressed that I’m going to make a mistake while driving, so 99% of the time, I just opt for shotgun.

Although the move towards driverless cars is nothing new, Honda just raised the bar. Their new system actually reacts to their surroundings and makes decisions off of that. Not that this should encourage anyone to play on their phone while driving, but it could add some security for people on the road. One of these Hondas could tell you it would be a good idea to switch lanes because the car in front of you is slow. If you approve, it’ll check if it’s safe for you to move, and then it will move lanes for you. That’s amazing! I don’t want to get too ahead of myself, but it’s like all of the issues I feel when driving would be solved.

2. AI is the newest addition in the quest to cure cancer

I’m fortunate enough to have been in good health for most of my young life, but many of my loved ones haven’t been as lucky. Breast cancer runs heavily in my family and sometimes it feels like it’s just a matter of time for me to get it, too.

AI is now being used for research on cancer treatment, and the current happenings seem optimistic. For example, Pfizer is using AI in drug development, including cancer targets. The AI sifts through and analyzes data from millions and millions of medical journals, to assist in the search for the gene that contains cancer so it can then be treated. So while a group of researchers could hypothetically read all these millions of journals that exist, it would literally take a lifetime or two. That’s a lot of lives lost to cancer, and AI is speeding up the process tremendously.

3. Fake porn… and Tom Cruise

Remember the fun Tom Cruise TikTok I included in last week’s piece? That’s a deepfake — an AI-generated video of a real person saying fake things. It can be fun and creative, like in this video of Salvador Dali. But if used maliciously, it can be really dangerous. AI Bots can, and do, create controversial fictional videos, blast them on social media, and call it news. If a story pops up on my grandfather’s Facebook feed that reinforces his political views, I don’t trust that he’d check the legitimacy before sharing. Facebook and Twitter have been battling this for years, but it’s so far from being solved. Deepfake software has been used to synthesize pornography with other people’s faces on it, ruining careers and lives. My photos, along with most people I know, are posted publicly on social media. It’s unnerving to think about how they can be manipulated and spread, and it’s hard to know if anyone can be safe from these threats.

4. GPT-3 might just signal the obsolescence of my career, so that’s cute.

Elon Musk has earned himself a lot of publicity recently, but it may not be for the right reasons. The latest software from his company, OpenAI, is the most intelligent AI model to date. It’s called GPT-3. It can write scripts, emails, poems, and music. If I wrote the first paragraph of this piece, GPT-3 could write the rest. GPT-3 isn’t perfect, but it’s a huge technological leap. It deserves a Spinning Towards Singularity piece of its own, because it’s extraordinary. It can synthesize graphics, sift through documents and write compelling narrative in a fraction of the time that I can. Once this develops further, many jobs will be easily replaceable. That indicates an upcoming cultural shift that will be totally detrimental for people who are unwilling to adapt.

Recent reads:

A brief roundup of articles that fascinate me.

This concept of synthesizing photos for the sake of “bringing beloved ancestors back to life” is, to me, equal parts sweet and strange.

https://www.theguardian.com/technology/2021/mar/01/deep-nostalgia-creepy-new-service-ai-animate-old-family-photos

I thought of this when I was writing about the threat of deepfakes and the security of my images online. Additionally, our mindless Instagram posts are being scanned to develop AI… so take that as you will.

https://www.cnbc.com/2021/03/04/facebook-trains-ai-to-see-using-1-billion-public-instagram-photos-.html

Final thoughts:

I try not to be too in-your-face about my dietary choices, but David Mintz, creator of the famous ice cream alternative Tofutti, just passed. And I think we need to pay respect to the unsung hero who made dairy-free living so much more attainable.

Have any questions? Send them my way. I’d love to chat — and if not, we’ll revisit next week. We’ll discuss my dog’s journey of learning how to not get his head stuck in the door.

Making Artificial Intelligence a little less intimidating