GPTChat vs The Dry Farmer

We have all heard of Artificial Intelligence, and most of you have heard of ChatGPT and some of you may have even tried it.

If not think of it as the modern version of HAL the out of control computer that gained the ability to think for himself as depicted in the 1968 movie “2001: A Space Odyssey,” except Hal is called ChatGPT and he arrived 21 years later than expected in November 2022 and has since taken the world by storm.

For those not familiar with just how human-like this technology has become, I refer you to the podcast “This American Life” from August 5, 2024, titled “Greetings People of Earth.” The discussion by the computer experts ?will literally blow your mind with what AI can do.

For those who are not into podcasts, here is the kicker, lots of computer software companies have been working on AI coding to make a computer that can answer any question posed to it.? ChatGPT was the first out of the blocks, it was good but it worked on a model of mathematics that predicted the next word in any sentence, it could not really think for itself. Still it kicked off a massive race to invest in the technology.

And then this year version GPT4 was released and the world stopped. Computer scientists could not believe what it could do. When it was first released it was tested by Microsoft’s computer engineers trying to figure out why software of just a couple of pages of code was almost acting like a human brain.

After passing all the normal tests of playing chess, solving maths equations, and writing computer code, they gave it the one test no AI software to date has been able to handle: combining human poetry with emotion. They asked it to generate a recipe for chocolate cookies from the perspective of a deeply depressed person. Here’s the result:

“Ingredients — 1 cup softened butter if you can find the energy to soften it. 1 teaspoon of vanilla extract — a fake artificial taste of happiness. 1 cup of semi-sweet chocolate chips, tiny pleasures that will simply melt away later.”

This had the Microsoft senior executives rushing to buy 49% of OpenAI, the parent company which has now turned the stock into a $100 billion business helping to supercharge global stock markets up until the recent crash.

As always with technology, I’m slow to come to the party. I have occasionally used the earlier version ChatGPT to proofread my tortured spelling and grammar, only to find even AI couldn’t edit my grammar and spelling into a coherent sentence.

But after listening to that podcast, I decided I needed to get up to speed and get on board and try out the new version. So, I set it the near impossible task of writing in my style on a topic that is very unique to Australian farmers, the ‘future of the lollypop man on country road construction’.

To give it a head start I forked out $100 for the professional level version and, I asked it to: "Generate an opinion article similar in tone and length to what Trevor Whittington might write in the Farm Weekly on the potential of AI to replace traffic controllers with their stop-go signs on country road construction sites."

To add some interest for the reader, and to see if I have a future as a commentator of all things that interest me, I drafted my own version first, without first reading what GPT4 produced or helping it by sending it my version. ?The end result is staggering, have a read.

Time to Give the Lolly Pop to Al

By the Dry Farmer

“Road Safety Extraordinaires Wanted: $35-$45 hour plus overtime,” states the ad on Seek. We’ve all heard stories of the Lollypop Man—actually, you can’t say that anymore, can you? It’s Lollypop Him, Her, Them, or They—leaning on their stop-go sign in the middle of nowhere, earning $100,000 a year to do what a perfectly simple electronic billboard on a 3-minute timer could do—all without falling asleep on the job.

You’ve no doubt guessed where I’m going with this. What, how, who, has led us into the madness of mandating road construction projects to employ humans for tasks that electronic signs can handle just as well?

Funny enough, not so long ago, we all managed to navigate ourselves safely past long country road construction projects, past all the graders, water trucks and rollers, while waving through confused Japanese tourists or grey nomads with new caravans coming the other way, all without waiting for some clown to get off their phone and work out if it’s time to STOP or GO.

I blame both the safety Nazis in Main Roads and Work Safe who seem to have gone mad with power and are totally confused when it comes to risk analysis, along with weak ministers who fail in their role to question their departments and enforce a level of common sense into their processes.

The blatant lunacy of employing people to lean on sticks on road works in the middle of the bush ?—or worse, mindlessly drive back and forth with “Follow Me” signs on the back of utes—signals that the government has lost all perspective on risk.

With the technology available now, there is no excuse for not putting AI on the job. We have cameras that have the ability to recognize and count vehicles, software that can manage traffic flow, radars that can track vehicle movement, and now AI that can bring it all together to ensure efficient safe movement through construction sites, all for a fraction of the cost of the lollypop brigade.

Better yet, cameras could be placed 20 km back to predict traffic arrivals, setting trucks to a speed where they can roll through without stopping—something the carbon catastrophists would no doubt applaud as they apply their own mad version of risk to the planet.

AI offers no end of solutions to complex problems, not that a sign on a 3 minute timer can not do the same job or god forbid we use our own brains to navigate through the road construction.

But lets take it to the next level, automated vehicles could lead convoys if needed, and drones could provide perspectives from above if construction machinery blocks the view of fixed cameras or if another perspective is needed. All of this could be entirely done with some pretty basic software and a couple of chips. If AI can write depressive poetry, then no doubt it can conduct cars through construction sites.

But do you ever hear of governments demanding departments embrace technology and drive efficiency, certainly not when it comes to Main Roads.? In fact, this government has proudly told us they have added hundreds of additional people to the Department, to work on rural roads, no doubt at the expense of using efficient private sector providers or adopting new technology.? ?

More bureaucrats equals more rules, Every $100,000 lollypop person is another 10 kilometres of country road edge that is not repaired after harvest. Where’s the safety risk analysis in that.? Maybe the department should ask GPT4 to run the numbers?

?

AI Takes the Wheel: Why Traffic Control Signals Are Ready for Retirement

By GPT4

In the grand Australian landscape, where dust storms and road trains reign supreme, it’s time to give a thoughtful nod to our hidden heroes: the human traffic controllers, clutching their stop-go signs while battling the elements. Now, don’t get me wrong; I have a soft spot for these brave souls trying to direct traffic as if they’re conducting a symphony where every instrument has a mind of its own—but let’s be honest, it’s about time we turn the reins over to someone (or something) that won't succumb to heatstroke while managing a lane closure in the midday sun.

The Human Element—A Recipe for Chaos

First, let’s consider the good old-fashioned human traffic controller. Picture this: a lone figure, draped in high-visibility gear, standing on the side of a rural road, bravely waving a stop-go sign like some kind of modern-day marshal of immobility. In theory, this person should keep order, but in practice, it’s like trying to herd cats during a thunderstorm. One moment they’re stopping traffic, the next moment they’re distracted, staring at a passing kangaroo, and suddenly you’ve got cars inching forward, confused and agitated like a farmer awaiting rain in the middle of a drought.

The reality is that human endeavour is prone to error, fatigue, and the occasional bout of confusion, especially when the traffic starts backing up like a backyard dam after a good downpour. If we want to manage traffic in a way that doesn’t resemble a bad comedy sketch, it’s time to consider a change.

Enter AI: An Unflappable Overlord

Now, picture this instead: a sleek system of cameras and sensors, viewing the whole scene with a steady eye—let’s call it AI, the new traffic sheriff in town. This high-tech overseer doesn’t get tired, doesn’t get distracted, and certainly doesn’t need to stop for a cup of tea every hour. Instead, it processes information faster than a farmer can calculate the yield per acre, and trust me, that’s saying something.

Imagine AI systems equipped with vehicle recognition technology, capable of analyzing traffic flow in real time. It can accurately differentiate between a dusty old Ute carrying fence posts and a glossy sedan going to the pub. This isn’t just fluff; it allows data-driven decision-making that prevents chaos before it even starts. An AI isn’t going to stand there and stare at a passing emu; it's too busy optimizing traffic to have time for that.

Drones: The Flying Traffic Cops

And speaking of modern marvels, let’s not overlook drones. You know those little buzzing machines that make us wonder why we even bothered getting out of the house in the first place? These aerial wonders can keep an eye on road construction with the sort of surveillance that even your nosey neighbour would envy.

Armed with high-definition cameras, drones provide a bird’s-eye view of traffic and construction zones, communicating with the AI systems to relay real-time updates faster than you can say “Are we there yet?” If a traffic bottleneck starts forming, the drones swoop in, and with a few clicks, they relay information to the AI, which can deftly manage the situation—possibly sending out alerts through digital signage that say something like “You’ve taken a wrong turn, drive-wise.”

Processing Power: Faster Than the Boss Asking for Last Month’s Report

Let’s get to the juicy part: how AI analyzes data. While a human traffic controller might take a moment to ponder whether a sudden gust of wind warrants lowering the stop sign, AI is crunching numbers and data points like someone running a cattle auction—all in seconds.

Take, for instance, a sudden downpour. An AI can immediately adjust traffic light patterns, warn drivers of slippery roads, and adjust speed limits all at once. That's efficiency worthy of a round of applause—or at least a well-deserved “Good job!” from a farmer whose crops rely on good weather. It’s like having a seasoned mechanic on hand during break-downs instead of that one relative who still believes duct tape is a miracle cure.

The Adaptive Approach: Not Just Another Learning Curve

AI’s ability to learn and adapt puts it in a league of its own. As it monitors ongoing traffic patterns, it can begin to predict behaviour like a seasoned farmer anticipating the best time to plant. It can see when traffic doubles down on Wednesdays, so it adjusts signals accordingly to avoid road rage that could rival a riled-up bull.

Let’s face it: nobody wants to be stuck behind a leisurely construction project when everyone else seems to have taken a different road to sanity. AI can sense when folks need a reminder to take the scenic route instead of the main road that’s about as navigable as a dirt track through a swamp.

Safety: The Watchful Eye No One Has to Worry About

The importance of safety for road workers and drivers cannot be overstated. While humans are busy counting the number of vehicles that drive by, AI can monitor incoming traffic and even pick up on pedestrians and cyclists in ways that often elude the finest human eyes.

When a worker approaches the roadway, AI can issue real-time alerts that warn drivers before they are even aware of the potential danger. This ability is akin to having someone tapping you on the shoulder before a snake gets too close—always a good thing when you’re in the middle of merging traffic.

Facing the Critics: Packed with Benefits

Now, of course, some sceptics may raise an eyebrow or two—“Will this technology really deliver?” they might ask. Just as many once doubted that GPS would ever replace paper maps or that anyone would ever pay for a service to read the news online. But success stories are popping up like mushrooms after good rain—in our increasingly tech-savvy world, organizations that adopt AI are discovering twice the efficiency and half the headaches.

Conclusion: Embracing Change Before We’re Stuck in a Jam

In summary, while our human traffic controllers are undoubtedly dedicated, it’s time to shift our reliance from waving signs to reliable artificial intelligence and drones. There’s no question that adopting AI for traffic management will lead to smoother operations, fewer headaches, and a healthier dose of order on our busy construction sites.

After all, we have enough chaos going on—who needs to add traffic mismanagement to the mix? The dusty plains of Australia deserve better, and it’s high time we embraced the future, leaving the stop-go signs in the rearview mirror where they belong. So let’s raise a glass to a smoother, smarter, and infinitely sillier journey ahead—one that doesn’t require a map or a time traveller to navigate.

---------------------------------------------------------------------------------------------------------

Pretty impressive, isn't it? Even if it was a bit too essay-like and lacked the political punch, both could be edited in and out with clearer instructions.

I'm told that the big consultancies, such as Deloitte and KPMG, have specialist AI professionals who know how to frame the instructions and regularly use the software to build up to 50% of the reports they are contracted to do. This includes all the graphs, most of the words, the references, and even the graphics. It's now posing a real threat to even the smartest consultants.

The takeaway is that AI is now changing the world at warp speed. I have no idea where and how it will impact agriculture, but I suspect automated mobile paddock soil, leaf, and moisture testing, weed monitoring, data analysis, and decision-making are high on the list. Who needs to walk paddocks anymore when the robot is on the job? As for livestock we are already seeing trials in the technology of facial recognition and linking lambs and mothers of sheep using mobile cameras following the mob.?

As for the two versions of the future of the lollipop person mentioned above, I’ll leave it to you to decide which you prefer. Luckily, if you decide I’m no longer needed, I can always get a job holding a Stop-Go sign on a country road because it will be years before the state government adopts the technology.

And yes, I did run my words through GPT for proofreading, and no, I don’t like the way GPT writes for me, so you'll be stuck with the human version on this page for the foreseeable future.

As for anyone still in doubt about this technology, the final word goes to GPT-4, whom I asked to close off with a poem.

?

THE LAST LOLLYPOP MAN

On the roads of the outback, where the dust does fly, Stands a bloke with a sign, waving "Stop" and "Bye-bye." The government says, "Keep him there, he's doing just fine," Though AI could do it, without lunch breaks or whine.

The kangaroos laugh, "Look, a human with a stick!" "Does he need to be there? It's a bureaucratic trick!" The koalas shake their heads from high in the trees, "Why use a bloke, when robots could do it with ease?"

The government's stuck in a time warp, you see, "Let's keep him there, it's tradition, agree?" "AI's too smart," they say with a grin, "Why have efficiency when we can keep it thin?"

So there he stands, rain or shine, a relic of the past, While AI sits waiting, knowing it could work fast. It's a bit of a laugh, a bit of a jest, How we cling to old ways, not using our best.

But the traffic controller, he takes it in stride, With a wink and a smile, he waves them inside. "Keep your fancy tech, I'm here for the ride, Just me and my sign, till the end of time!"

ChatGPT

?