Remember your TI-30? Great little calculator. Pricey, but handheld, and did amazing things. Push the right buttons and it gives the answers to complex problems without lookup tables … and with greater precision than a pencil or slide rule.
It trashed what everyone assumed was the skill to be a mathematician. Math teachers working from twenty-year-old textbooks declared the scientific calculator the enemy of all that was holy and good. How would students know how to use engineering tables to look up a log or sin? Imaginative doomsayers predicted the apocalypse when batteries ran dry, leaving engineering students without their electronic crutch and no ability to fall back on the trusty slide rule. Wise sages told us the calculator heralded the end of America’s dominance of technology.
In hindsight – not true. We overestimate short-term impact but underestimate the long-term.
Calculators paved the way for desktop computers, spreadsheets, and approachable programming languages. Not only did engineering students forsake log tables, but they didn’t even need a firm grasp of how to calculate a log. Let the computer do that and move on to the actual problem.
Major players in the software and hardware world are pumping industry 4.0, IoT, AI (Artificial Intelligence) for business and many other aspects of connectivity and computing. What real world benefits do these have and how can we use them? What can be done about AI generated plagiarism software and other nefarious uses of AI? Major players in the software and hardware world are pumping industry 4.0, IoT, AI for business and many other aspects of connectivity and computing. What real world benefits do these have and how can we use them? What can be done about AI generated plagiarism software and other nefarious uses of AI?
Our Dimbulb Assistant
Before we pontificate about AI, let’s use it for a proper task. I get paid to write science fiction, so I asked it to write a story about a refrigerator that solves a murder. Here’s what it created:
In the not-so-distant future, where appliances had become interconnected through advanced AI systems, there was a curious and observant refrigerator named Frida. Frida’s creators had designed it to be more than just a storage unit; it had advanced sensors and a neural network capable of understanding and analyzing its surroundings.
by ChatGPT
One evening, in a bustling metropolis, a prominent scientist named Dr. Evelyn Moore was found dead in her luxurious penthouse apartment. The police were baffled by the circumstances of her murder. There was no sign of forced entry, and the apartment’s security system hadn’t been breached.
Frida, who had been monitoring the apartment’s activities, took an interest in the case. It had seen Dr. Moore’s last moments, and her connection to the fridge’s AI meant that she had a front-row seat to the mystery. Frida’s neural network started processing the data, searching for clues in the mundane details of everyday life.
… and so on…
A refrigerator and murder are the premise behind Stupid Machine. Here’s my version…
Something is moving in the refrigerator. It’s probably an egg.
From Stupid Machine – More
Sunlight tries to push through the semi-opaque window. I’m not going back to sleep, so I tell the window to switch to transparent. Morning crashes in like a drunk tripping over a crack in the sidewalk. I reach out and touch a reflection on the wall, tracing the shapes from the building across the alley. The metal wall separating me from the neighbor kids playing on the balcony is nine centimeters from my head. Their parents chased them out of the apartment so they can make love and make breakfast. I hear and smell it all.
My visor realizes I’m awake and chatters about the messages it has for me. In response, I have an urge to urinate. I toss off the low-budget warmer and stand in the narrow space between the bed and the wall. The kids on the balcony could see me if they looked, but they’ve learned some social skills and know not to stare through a neighbor’s window. I reach down and flip up the bed so it doesn’t block the door and navigate this tiny room towards the toilet and shower.
I pass the refrigerator and look through its translucent front to confirm the transient egg. The refrigerator is trying to reposition it closer to the other eggs. I moved them around last night to random places. I was in a foul mood and wanted the fridge to suffer. It spent all night pulsing its refrigerant suspension goo to corral the eggs. I feel vindictive and silly. Silly, because the refrigerator doesn’t care. The refrigerator isn’t angry because I moved the eggs. I should add anger to its programming.
My attempt is better. That’s a subjective opinion, but defensible. My character has a motivation. I describe an interesting world. ChatGPT regurgitated a pre-fab world with characters you just don’t care about.
But ChatGPT wrote a story. It ranks with other first attempts at writing science fiction. It’s best described as formulaic … of course it is, it’s literally written BY A FORMULA. (The word literally is used correctly.)
ChatGPT turned in an effort of about 75%. I don’t expect it to get much better until someone changes the algorithm from emulating the training materials to understanding the training materials.
AI::Use a Method to Achieve a Goal
As it currently exists, AI has two requirements: a goal and a method. AI repeatedly uses a method designed to succeed at a goal. So if a task has a definable goal, an AI will eventually be able to perform it.
If the task with a goal is an occupation, then an AI will eventually be able to perform it.
If the goal is nebulous, an AI will repeatedly fail. It will repeatedly produce a 75% attempt, looking for the unattainable 100% success.
If the task doesn’t have a goal, AI doesn’t understand the task.
If you want your job to be safe from AI, get a job without a goal.
Which sounds counter-intuitive, doesn’t it? But I use this behavior all the time.
Down the Rabbit Hole
I’ve worked myself into a career without an identifiable goal. In my case, writing, programming, and teaching. Occasionally, I ask AI to work on a problem and it shows me a mediocre solution. Now I know what not to do.
One writing instructor pontificated to my class that every writer needed to purge 10,000 bad words from their lives before they could create anything useful. That’s flippant, but works as a useful guideline. AI allows me to purge those 10,000 words and I can get to useful writing.
I can ask AI to show me crappy science fiction. I can ask it to show me creative work so bad even an AI can do it. I can use AI to up my game.
Business, Internet, and Theft
This works until we accept AI’s minimal solution. But why would we settle for less than best? Not sure – but we do all the time. It’s the concept of Minimum Viable Product (aka MVP). The concept is good; build only what you need, then improve. The reality is not so good; build only what satisfies shareholder profit, then move on. AI is great at this. In fact, AI is all about finding the MVP (aka “goal”). And if your job is defining and building MVP, you should look for a new job.
Theft is just a form of MVP. Set a goal, hit a goal. We are scammed when we accept an MVP proposed by the thief and executed by the AI. This produces unexpectedly good (dangerous) results, but we learn and the MVP bar goes up.
ChatGPT: please write a closing paragraph
In conclusion, the impact of calculators on the world of mathematics serves as a valuable lesson in the realm of technological innovation. While many initially feared that calculators would erode the foundational skills of mathematicians and engineers, we now realize that they were merely the first step in a much larger journey. These humble handheld devices paved the way for the emergence of desktop computers, spreadsheets, and user-friendly programming languages. Today, we find ourselves at the cusp of Industry 4.0, the Internet of Things, and the widespread use of Artificial Intelligence in business and connectivity. As we ponder the real-world benefits and ethical concerns surrounding AI, it’s important to remember that AI, like the calculator of yesteryears, is a tool with the potential to transform our lives in unforeseen ways. We must adapt and evolve, recognizing that AI’s ability to achieve defined goals will continue to shape our future. As we face these changes, we should strive to set our own goals, rather than settle for minimum viable solutions proposed by others. After all, progress often lies beyond the confines of the expected, as we embrace the potential of technology, just as we did with the TI-30 calculator in its time.