As artificial intelligence continues to improve, we humans will have to decide what we are comfortable letting it do.
What would it be like when you had your own bot and it was as good or better than you at many everyday tasks?
The answer may come sooner than you think. Google’s new Language Pathway model, which has yet to be opened for public testing, is the latest advancement in artificial intelligence. The technical explanation is that the neural network has been expanded to 540 billion parameters for “breakthrough performance”. The practical effect is that AI is now better at engaging in natural conversations, interpreting novel jokes, and coding.
I expect most of the written communication will eventually be done by bots. I can train my bot by having it read all my past emails and other posts. Ultimately, my bot will respond to most of my emails directly, although it may keep some emails aside to ask me if they deserve a personal response.
This sounds convenient, and in many ways it will be. I will have more time to go for a walk and read a book. But think about the broader equilibrium. If more emails are read by bots, then more emails will be written by bots. Of course it is, but in this new world bot-drafted emails will be at least as good as human emails and at least as good at bypassing whatever filters I set up to protect time. my time and attention.
An arms race will follow. Overall, I expect the volume and quality of emails to increase. Woe to those who don’t have very good filtering bots.
Imagine negotiating or discussing terms in such a world. I can get a recommendation from your bot. Is it a real, legally binding offer? Or was it simply a ploy to get me to disclose information about my negotiating strategy? In some cases, bots can handle these issues smoothly and offer a final solution for both parties. In other cases, negotiators may demand a face-to-face meeting to know that they are reaching a “real deal” and to limit the possibility of back-and-forth. For some real-world interactions, online written communications won’t be good enough anymore.
For example, think about the college admissions essay. Today it is important. But if the bots turn out to be well-written, the candidate might have to come in for a personal interview instead. Countermeasures may then develop. There may not be enough admissions officers to conduct all of those interviews. So why not let the candidates spend two days together, document all the procedures and let the bots come up with the ratings? They can even measure who tells the most original jokes.
In this new world, writing skills will be much less, and personal charm much more. This is not necessarily a positive development. It is more difficult to use the written word as a measure of skill or broader intelligence.
If you’re single, bots could change the way you use dating services. Does it seem tedious to have to swipe left or right all the time – and besides, do you really trust your own judgment? Instead, you can let your bot choose for you. If you tell her you’re interested in a potential mate, she might even send you pictures of what your kids look like.
However, as with email, there are some potential complications. The good news is that your bot can sort through the entire pool of available candidates quickly. Maybe your soulmate was at swipe level 3,472 – and maybe now, thanks to your bot, you find that person instead of giving up. The bad news is that the candidates who are the easiest to marry and the easiest to marry may be left out of the group of the most beautiful. The liquidity of the dating market can dry up for less desirable candidates. Tinder obscures you from the fact that a candidate might just be a 4 on a scale of 1 to 10, but your bot is not.
Artificial intelligence is of great benefit to many tasks, such as detecting payment fraud, improving medical diagnostics, and sending rockets into space. Potential problems arise when AI systems interact directly with human attention — and the activity involved includes a lot of matching and filtering. In those cases, AI advancements may overwhelm the ability of humans to participate in the process. And there is probably no final stage, at least not soon, where we can fully rely on AI.
