Mad About Science: Artificial Intelligence I: The Present

By Brenden Bobby
Reader Columnist

If you’ve made it this far and didn’t scoff at the title, welcome to the present! After last week’s antics we know the basics of how AI works. You input constants, it runs them through algorithms and cross-references them over past results to tailor future results to your personality and desires. Kind of like how your grandma gets you slippers every year because you liked those slippers she got you when you were five, but with the internet as grandma’s brain.

AI comes in several distinct forms, but the kind we interact with the most today is called weak AI, or narrow AI.

This type of AI exists to execute a specific function. Search engine results, most notably. Personalized ads on those search engines.

Predictive text, depending on your phone’s operating system, is one of the most interesting AI developments we’re experiencing now, in my opinion; but hey, I’m a word nerd so what do I know? Predictive text and autocorrect (artificial unintelligence, am I right?) will record your typing habits and offer suggestions or straight-up edits based on how you’ve talked to people in the past. Human speech is an incredibly nuanced thing, and the fact that my personal style can be predicted by a machine that’s watched me complain about bad drivers is awesome to me.

Similar to predictive text, AI’s ability to adapt and learn is being applied to finances, from small business to giant mega-corporations for sniffing out errors or financial misdeeds.

Search engine algorithms are the biggest way weak AI influences our lives now. Companies like Amazon, Google and Netflix almost know you better than you know yourself. The computer can go as far as deducing what you enjoyed more than other items based on how much you searched for it, how long you searched for it and how much you consumed, then making predictions and guiding your future interests based off that. In a way, you’re allowing the computer to manipulate your tastes.

This kind of AI gets even freakier. If you have kids that have spent as little as an hour on Youtube, you’ve probably been subject to an avalanche of questionable knock-off videos featuring popular animated characters. Why is “Peppa Pig” getting blown up by “The Avengers” on the back of “Garfield”?

Many of these videos are actually created from start to finish by AI, mashing seemingly random characters and audio files together based entirely on their search popularity. This may seem like a completely frivolous exercise to an adult, but it’s causing an immense stir among parental groups online. Between the mix of random gratuitous violence the computer sometimes picks out and the nonsensical brain-melting gibberish the computer decided to make Thanos say, these videos are shaping the tastes and interests of children in ways we’ve never seen before. Sure, “Scooby-Doo” probably influenced some of your more annoying go-to lines as a kid, but the show was overseen by humans, approved by humans, made sense because of humans. Now we’re talking about a computer picking everything based entirely on search popularity.

Not to be alarmist, but what impact does that have on our society or our culture when an entire generation is raised on that?

Other relatable, less-apocalypse-y AI uses in our lives come in the form of personal assistants: Siri, Alexa, Google, Cortana, et cetera. These personal assistants use multiple layers of AI, from learning how to decipher your voice to communicating with satellites based on your location to find your most relevant results. The sheer power of the phones in our pockets is staggering, yet it’s basically just an extra appendage for us now. It’s become so normalized by use that we don’t even think about having HAL 9000 in our pocket.

Self-driving cars are at the cutting-edge of AI research as well. We’re just dipping into this now, so there’s no way we can even comprehend the totality of implications this kind of invention can have on our lives. We’ve seen human fatalities can still occur, so they’re not infallible, but does that really mean we need to completely stop researching them? What sort of implications might that have on our careers if our employers decide you can get an extra 40 minutes of work done on your commute? What if humans can never drive again?

Healthcare is perhaps one of the biggest beneficiaries of artificial intelligence. AI is capable of aiding your health from the moment you step into an office to the end of an arduous, but successful cancer treatment. AI has been developed to comb through health records, cancer records, drug trials and much more to present effective treatment options to your doctor tailored to your body. We’re not talking WebMD here, we’re talking comprehensive, AI-crafted treatment plans from start to finish with shockingly effective track records. Doctors have even begun to apply AI-controlled robots to complete difficult surgeries that require precision a human hand may lack.

This is the snowflake at the tip of the iceberg when it comes to AI we’re using right now. Next week, I’ll talk about what could lie in store for us in the future, and why that does a fantastic job at scaring the pants off us and redefining us as a species.

Hasta la vista, baby.