The Bots are Coming

Unless you’ve been living under a rock, you’ve heard about new AI tools like ChatGPT and their amazing abilities to imitate humanity. I mean, for all you know, this could have been written by an AI. (It wasn’t, AI would probably have far fewer run-on sentences).

The thing about humans is that we’re remarkably good at detecting things that aren’t quite human. We call it the “uncanny valley,” this uncomfortable place when things seem too human to be animated, yet too animated to be human. And as humans, we’re not comfortable yet with interacting with AI. We think it’s cool, but as interactions with Bing’s Sydney demonstrate, things can go sideways quickly, and in very unpredictable ways. 

AI researchers have understood for quite some time that large language models have interesting issues around prediction and some can even become self-referential. There are some fascinating cases where a lawyer is being sued and may even be disbarred for relying on ChatGPT — it made up the cases that it referred to in the legal document he had it draft as precedence — none of them existed! Deep-fakes are a major concern, calling into question how we will understand what is real, and what is not real in the future.

We’re also now noticing that AI art is starting to imitate itself, leading to questions about the true limitations in the generative space, which some critics did predict, as natural language models are built around imitating previous content — it cannot imagine therefore it cannot innovate.

Software companies are turning to AI in droves, each one rushing to implement it faster than the other. Medical, legal, and other professional disruptors are popping up faster than mushrooms utilizing AI to replace the human beings that have historically been the ones to review your cases and determine what is wrong with you. The problems with these models in these use cases are many — biases in AI can be implicit, outliers are often not considered, and diagnoses can be missed. 

Another place where AI is promising disruption is support — we’ve all had our fair share of frustration with automated telephone systems but now the bots are threatening to replace our support workers too. How will they act? Will they truly be able to understand your business needs or will they be stuck only understanding a piece of software? As good as Alexa and Siri are, they still struggle to turn on and off the correct lights at my parent’s house. I have my doubts. 

Every day here at Estimate Rocket we hear from customers that are thankful they can reach a real human on the phone and that our folks on the chat are real live humans who actually know a great deal about our software and your business. And now we’re hearing that the tech industry at large doesn’t even want you to speak to your doctor or to legal experts? 

That’s ridiculous.

Our deepest purpose — that we seek to truly solve your problems, not just your technical ones, but the hard problems that plague your business — relies on people who understand your struggles each and every day.

When you have a problem, do you want to speak to someone who can have true empathy? Or do you want to speak to a bot who may have a limited ability to translate or really get deep into the issue at hand? 

Your choice.

Does that mean that we aren’t investigating AI? No, not at all. But we’re very interested in using it responsibly. We have been investigating what this technology means for you so that we can not only use it in ways that work to function for your business but also in ways that you can rely on. 

We never want to just leverage the new and cool so that we can sell more software. That would be totally against our core values.

Our business is centered around giving you exactly what you need so that you can count on your books being right, your customers being happy, and not paying for things you don’t need because we think it’s shiny and fun. We aren’t tech bros hanging out in Silicon Valley, we don’t have those fun cafeterias, and we definitely don’t play ping pong all day or whatever you’ve seen on TV. Our dedication is to your business and to your success.

Back to AI — do I believe that it could be helpful ultimately? Yes. But rushing to implement it without fully understanding the implications could be doing you a massive disservice. As a society I’m not sure anyone is really talking about the systemic issues it could cause. Sure, there are a few people raising alarms, but no one is really talking about them seriously – yet. My hope is that we start to have some serious conversations before it’s too late and we find out the hard way what the implications are for things like truth and creativity.