2023: The Year AI Kinda Blew Up (Except It Didn’t)
- By Winston Thomas
- January 01, 2024
Remember Y2K? Are the robots rising? 2023 promised the same tech-apocalypse vibes, but instead, it threw us a curveball so gentle it was like getting whacked with a feather pillow. We expected laser battles and cyborg uprisings but got... spreadsheets on climate risk and ethical debates about ChatGPT's existential angst.
Don't get us wrong, the year was a heady cocktail of geopolitical jitters, economic hangovers, and enough AI drama to fill a dozen cyberpunk novels. But instead of robots taking over the mines (yet), they were busy making them safer (because apparently cyborg miners are bad for PR). And while everyone was freaking out about AI poisoning data and spewing half-truths (seems ChatGPT's got more baggage than a Kardashian), the real story was how this tech was quietly making the web more accessible for people with disabilities. Talk about a plot twist.
Our discerning readers lapped up all this AI goodness. Here are the four of our top stories (in no ranking order) that tickled their techy taste buds the most:
When AI gets on acid
In his article "The Data Conundrum of AI", Paul Mah took a deep dive into the murky world of data poisoning, where bad guys can tweak public data to make your AI models trip harder than a teenager at a techno rave. Turns out, ChatGPT's existential crisis might not be about its robot feelings, but the fact it just ate a bad batch of web data.
Mah also quickly pondered the dilemma of half-truths in ChatGPT replies. He swallowed his pride and asked the model to tell more about himself. The spewed replies were neither shocking confessions nor heartwarming truths. It was like prying open a fortune cookie to find a stale proverb about the dangers of small data.
Another find is that the ChatGPT is very helpful for data scientists to manipulate data—but only at an infinitesimal scale of one record. Want to nudge a number here, massage a value there? ChatGPT's your pal. But go beyond that microscopic level, and it starts making things up to give you more than the "truth" you bargained for.
Moral of the story? Fact-check your AI like you fact-check your Tinder dates.
When climate risk becomes a spreadsheet nightmare
In her article "Surviving Future Storms With More Data", Sheila Lam painted a dramatic picture of climate change, not as melting glaciers and angry polar bears, but as a mountain of compliance paperwork for companies.
Apparently, exchanges are now asking for climate-risk data like it's the latest cryptocurrency (spoiler alert: it's not). This means data teams everywhere are scrambling to brush up on their climate-risk vocab, all to avoid drowning in a sea of fines and PR disasters.
But fear not, intrepid data warriors! Lam identifies the two krakens in this stormy sea: internal data collection that's more Frankenstein's monster than a well-oiled machine and external sources that are about as reliable as a used car salesman on the moon.
Her conclusion? Every company, not just publicly traded behemoths, must get their climate data act together. Investors are sniffing out greenwashing like bloodhounds on a sustainability scent trail, and nobody wants a future powered by coal and regret (at least in plain sight).
Making the Web less of a jerk
While everyone's been gushing about ChatGPT's productivity superpowers, I took a different angle in the article “We Are About To Thank ChatGPT for Making Web Accessible”.
Apparently, this AI whiz kid can also make the web more inclusive for people with disabilities. Imagine websites that actually work for everyone? Mind. Blown.
This isn't just about feel-good vibes; it's about opening the digital world to a new audience. So, the next time you hear AI doomsayers ranting, remind them that robots can actually make the world a better place (as long as they don't start writing your emails).
Of course, GenAI is no accessibility saint. It trips over its own code like a toddler in a mosh pit, spitting out hallucinations and unpredictable results that make you question if you're lost in the internet or the Matrix. But hey, even a glitchy AI hand is better than no hand, right?
Think of GenAI as training wheels for accessibility. Web users can get a taste of what a barrier-free web could be, while designers can see where their interfaces are tripping up even the most tech-savvy surfers. As one interviewee wisely pointed out, GenAI's kinks will get ironed out eventually. It'll get smarter, more nuanced, and churn out results smoother than a freshly paved digital highway.
Let's just hope those robot glitches don't lead us down a dark alley of autocorrect fails and existential errors.
Robots invade Aussie mines (but don't worry, they're friendly)
Lachlan Colquhoun took us on a walkabout in the Australian outback in his article “Robot Miners Take Over Western Australia”.
In his words, he described Australia's future: robot-run mines where machines do the dirty work and humans get to play office jockey. These robotic miners, we find, are surprisingly chill, making the mines safer and more efficient (and probably less prone to union strikes). Who knew the future of mining would involve spreadsheets and autonomous trains? Not us, that's for sure.
The first stop in Lachlan's digital tale was the Roy Hill iron ore mine, where autonomous haulage systems glide around like driverless drones, boosting efficiency and ensuring nobody gets squished by a rogue dump truck. Then, he throws us onto a high-speed train (well, a metaphorical one) to Rio Tinto, where they're building the heaviest-duty railway on this side of the planet.
But this ain't just a robot rodeo, folks. Colquhoun paints a bigger picture: AI and alternative energy are the new gold dust. He throws around terms like "digital twins" (basically, a virtual clone of your mine you can tweak and prod without getting your boots dirty) and "large-scale solar farms" (imagine powering your entire operation with sunshine).
So, remember Colquhoun's vision the next time you see a lump of coal. The future of mining ain't just about digging deeper—it's about digging smarter, with robots, AI, and enough clean energy to make Greta Thunberg do a victory dance. Just hold onto your hard hats, because this robo-revolution is coming faster than a runaway haulage truck on a downhill slope.
2024: Buckle up for a wild ride
So, 2023 wasn't the AI apocalypse everyone predicted, but it was still a year that kept us on the edge of our virtual seats.
2024? Buckle up, buttercup, because it's gonna be a wild ride. AI's gone mainstream, robots are digging for gold, and the world's finally waking up to the ethical can of worms this tech unleashes.
We've got U.S. elections, Asian election fiestas, conflict zones galore, and enough AI exploits to fill a black market bazaar. One thing's for sure: 2024 will be predictably unpredictable.
Happy New Year, you bunch of tech-savvy humans!
Image credit: iStockphoto/Brosko