I need an adult

 I've explored some of my philosophy on artificial intelligence on this blog a time or two. And I continue to be very conflicted about AI. 

Oh, the negatives. We've all heard them. Overuse of our already strained resources, concentration of power and control in the tech oligarchs, concern about my own security and autonomy, loss of skill and concerns about the quality of the information. Not to mention the absolute tidal wave of slop covering basically everything in the digital world. 

But there are some positives we need to recognize too. I made good use of AI-adjacent tools when I was off injured this summer, and I've made the argument in the past that it could be a real boon for people with differing needs. And it's a productivity tool - kind of an unsettling one. And that's where we truly begin today's post. 

There's been steady pressure at work to use Microsoft Copilot to help with the day-to-day. I haven't been able to really throw myself into it that much. I've used it to help me break a bit of writer's block- to make it start a news release or a speech for me. But inevitably, I end up rewriting it. It's like how Sherlock Holmes enjoyed having Watson around, if only to show him the wrong answer so he could find the correct one. Using stupid to make you look smarter. Ol' Sherlock was way ahead of his time. 

But last week I had a conversation with my manager, and we wondered if we could use Copilot to help us make the morning news-scanning job a bit more efficient. 

So, dutifully on Thursday afternoon, I cranked up the ol' Copilot machine and started asking questions. 

This is where I need an adult. 

In some ways, it did what I'd expect. It answered the question I had using online resources and its access to our company documents. This is a promised feature, and it's actually pretty powerful. It used Microsoft documentation and forum posts to highlight what was possible and included references to our own procedure documents on how it would ensure that what we built followed those processes. Kinda cool.

But the unnerving part is that it was also pulling from my conversations on Teams and using those discussions to gauge my feelings on the subject and how others on my team might respond or what they might ask for. Genuinely creepy. Like, just having your conversations followed feels uncomfortable, but having a machine analyze those discussions and come to a conclusion about feelings and thoughts is really creepy. 

I suppose this is where I'm entering new territory. When I've used free/publicly available AI tools, they've only had the public internet to draw upon. Really just a very fancy Google. But giving that same AI access to my work and personal documents can really open up what it's capable of - for better or worse. 

I dunno. I continue to wonder if I'm being a Luddite or if my concerns are legitimate. I suppose the fact that I'm interacting with the technology, asking questions and expressing concerns based on experience means I'm at least better than a Luddite. They'd just say no. 

And with this new phase of discovery, has come conversation. And with that conversation, my algorithm has recognized the increased discussion of AI. With that recognition has come a wave of ads for AI tools. Annoying for sure, but those ads also highlight more of the things a person can do with AI. This has the knock-on effect of making me think about how else I can use it - at work and at home. But how much do I want to let this into my life? I could add Gemini to my existing Google subscription for only a couple bucks more a year. But I don't really need it highlighting a dirty text I wrote to Rhonda. Or try to buy that lap steel I keep looking at on Facebook. 

I still really don't know. I'm uncomfortable, but I see a world of possibility. It's like walking into an orgy you really didn't expect - but here it is. 

And they're beckoning you in.