So there’s this new AI platform called ChatGPT, right? And when it launched back in November 2022, it also came with an API to GPT. Well, not to be outdone, Google came out with their own API key for their rival AI platform Bard.
Now, the result of all this is that AI cloud services are poppin’ up like weeds in a garden, man.
Check this out, these AI cloud services have different tiers, like the standard free tier, the midlevel tier, and the premium tier that costs some dough. The thing is, as IT managers, we gotta be careful ’cause some of these AI solutions might not actually bring any real value, they might just be distractions.
Now, tiered pricing for cloud services ain’t anything new. It’s been around for a while, you know? Back in the early days of the web, people would mainly download software, so providers started offering different tiers to cater to different customers. Like, if you only needed the software occasionally, you could go for the free or low-cost option. But if you were a frequent user, you’d have to cough up more money to get access to more features.
So, what’s happening now with these AI tools is kinda similar. Some of them have pricing that’s in line with the old tiered models, but some are takin’ it to a whole new level. Like, there’s this one tool with a $19/month tier, then it jumps all the way to $149/month, and then there’s a completely customized quoted tier. It’s like every tier is a smaller increment, from $19 to $39 or somethin’ like that.
But here’s the thing, training AI models ain’t cheap, man.
These large language models need huge databases to be trained, and the cost of actually training these models has skyrocketed. Like, according to this report from Stanford, the training cost for GPT-2 was $50,000 when it was released in 2019. But now, the training costs for some models are reachin’ up to $8 million. That’s a huge difference, man.
So, these solution providers, they gotta cover their costs somehow, right? And one way they’re doin’ it is by incorporatin’ those costs into the pricing of their software. Even big tech companies like Salesforce are raisin’ their prices because of the economic factors and the implementation of AI in their tools.
But the price hikes aren’t just happenin’ to big companies, man. I’ve seen it happen to smaller products too. Like, this app called Evernote, I’ve been usin’ it for years. They recently announced a price increase, promisin’ new AI-enhanced features. But here’s the kicker, I can’t even use those new features ’cause they’re not available yet. So, I’m basically payin’ for AI to come, instead of gettin’ actual AI features that I can use right now.
Now, don’t get me wrong, some of these AI tools are actually pretty cool. They can really enhance productivity and make things easier. But here’s the catch, man. They can also create what we call technical debt. It’s like a trade-off between launchin’ a simple version of a product versus a version with all the desired features. Sometimes, it’s easier and cheaper to just go with the simple version, even if it’s not gonna meet all your strategic objectives.
And that’s the risk with these AI-based solutions, you know? They might automate processes and make things quicker, but they don’t always deliver the results you expect. And over time, people lose interest in ’em and don’t wanna spend more on maintenance and stuff.
And it’s not just about productivity, man. Even marketers are gettin’ caught up in the AI frenzy. They’re usin’ AI tools to create content, but it can create a whole other problem called “content debt.” It’s like a flood of content that needs extensive rework to fit different types of media. And that takes up resources, man.
So, how can IT managers handle all this?
Well, one thing they can do is conduct IT audits to see how different departments are usin’ these AI tools. ‘Cause it’s not about individual usage, it’s about how many people are usin’ the same expensive software. You don’t wanna be payin’ for unnecessary subscriptions, right?
Another thing IT teams can do is keep an eye on new AI developments. There are more open source frameworks and models bein’ launched, so teams might start developin’ their own solutions. And that’s gonna add to the costs, my friend.
Look, AI is definitely the future, man. It’s gonna keep growin’ and it’s gonna keep changin’ the game for a lot of industries. But we gotta be smart about it. We gotta make sure that the costs are worth it and that we’re gettin’ real value from these AI solutions. AI can empower creativity and productivity, but it’s up to us IT managers to find that balance.