Feb 25, 2023
Will Your Next Chat Move Your Cheese?
Those of you who have been reading my editorials since 1998 know I’m hesitant to adopt new technology without obsessing over it first. I must research and fret before installing an app that uses up precious data on my phone.
There may come a day when artificial intelligence (A.I.) helps me with some portion of my job, but I can make a promise to readers right now: the content of AsphaltPro Magazine is generated, curated and edited by human beings who know, respect and understand the industry. Although I read the same business book from Spencer Johnson that everyone else did in the ’90s, I’m content to sit back and watch how A.I. learns and grows before I log in and play with it for even personal purposes.
You might be wondering how my reluctance to trust a new machine civilization relates to the asphalt industry. Let me see if I can easily reveal the mental gymnastics I performed when applying ChatGPT to telematics.
Buckle up for this one.
Now this conversation applies to more than equipment telematics. Consider the volume of data you traffic in, as a member of the asphalt industry.
From tonnage delivered to a paving site per day to reclaimed asphalt pavement (RAP) fractionated for Pile 0.5-inch versus Pile 0.25-inch, you’re tracking materials. From liquid asphalt cement (AC) graded for FOB customers to the pricier AC with an additive blended at the terminal for a state job your crews will start next week, you’re tracking (and monitoring temperatures for) incoming perishables that have a serious price tag attached.
Depending on your platform-of-choice, all your data could be stored on hard drives that automatically back up on a predefined schedule, a third-party server, the omnipresent cloud, or some combination thereof. If you’re lucky, you’ve got it all going to the same place to be accumulated, evaluated, extrapolated, manipulated, and returned to you in some format that makes it easy for you to see what you have, what you’ve used, how you’ve used it, when it’s time for routine maintenance, when it’s time to order more AC, and what data points line up to make some predictions for next season. If you’re lucky, the entities performing this service for you are worth the cost.
What if an entity similar to ChatGPT’s openai could do it all for free? Has anyone in the construction industry tried feeding all their data to an A.I. platform to let the machine overlords do the work? I’m not suggesting it’ll be easy at first. But then, setting up all those touch points to feed to your current data-gathering mechanism wasn’t easy, either.
Instead of directing your tank temperature gauge, plant alarms, wheel loader’s maintenance schedule and slat conveyor’s wear thickness measurement to a couple different platforms, are we almost at the stage where we direct all those data points to an A.I. that accumulates, evaluates, extrapolates, manipulates and returns projections based on your complete data? For free?
I’m fully aware that the editors at CNET came under fire for using A.I. to generate financial articles that were riddled with basic errors. I’m fully aware that the Stargate A.I. writes scripts that have silly mistakes (such as having characters suggest “resting” when action is required to propel the plot). So, we’re not at a point where handing over all the telematics (and other) construction-related data could reliably offer sound business advice. Or are we? Is the next chat about A.I. changing someone’s business plan?
Stay Safe Out There,