Did you find the fake pages in Tech May?
A difficult concept to fully understand AI quietly slipped into the background of our online shopping, music listening and video streaming services. And then one day in November 2022 it was suddenly ready for prime time, and the text AI tool ChatGPT was opened to the public internet.
ChatGPT quickly caught the imagination of the general public and awareness grew quickly, especially as its ease of use and genuine usefulness proved itself in everyday life.
But what if it hadn’t caught on? What would happen if its launch had gone unnoticed in the news cycle, overshadowed by stories about crime, global warfare, loadshedding and maybe a dodgy Cassper Nyovest tweet?
That is where we are coming from in our May issue of Tech Magazine. If we didn’t tell you about it, would you know if the magazine was being made entirely by computers? Could you be fooled? We did an experiment placing computer-written text throughout our May issue of the magazine. Did you spot the AI-written stories?
HOW IT WORKS
Extremely powerful supercomputers are given billions of pieces of data, numbers, pictures and text, and then given a small mathematical calculation called an algorithm, a sort of mini program, to help it figure out how these pieces fit together.
The most famous AI, called ChatGPT, was allowed to trawl through all the information on the internet and then try to figure out the relationship between all the facts, figures, names and places it found. Of course, supercomputers have been doing that for a long time already, but now the AI knows how to create entirely new stories out of the information it has collected. This is why we refer to it as “intelligent”: it can create new and unique stories entirely by itself.
Just to be clear: it is not simply cutting and pasting bits of web pages together. It is creating completely unique stories from the information it has collected, and can place that into sentences, lists and formats humans can understand.
HOW WE USED IT
ChatGPT is the world’s leading text creation AI. It works with a very simple web site where you type a description of what you want written, and within seconds your newly created text appears on screen, and you can copy and paste it to use on to your PC.
WHAT WE DID (click on the thumbnails to read the articles)
On the ChatGPT web page we typed “write 300 words about Chat-GPT ” and here is what ChatGPT produced for the magazine, published on page 14. Turn over the page to see the original story about ChatGPT, written by our (human) technical writer Lindsey Schutters.
We typed “write a 300 word overview of key considerations when buying a home security camera” and the result was published on page 20.
We typed “write a 200 word review of the Civilization VI game for Android” and the results are on page 53.
We typed “write a 300-word editor’s letter for a consumer tech magazine about AI and chat-gpt” and the results are on page 3. Next to that is the human-written Editor’s Letter which it replaced. Which do you prefer?
Ed’s Letter, May 2003
Flicking through my Spotify dashboard, which is almost as busy as my Outlook dashboard, I encountered a playlist, three levels deep, which is a mix of classical music shaped just for me. Out of pure curiosity I clicked… and discovered the most perfect playlist of music I could have dreamed up.
All my favourites were there mashed up with some pieces I had always meant to get around to, and some pop music with classical music riffs that somehow blend seamlessly into the mood and tempo of the classical music. It was a playlist that would have taken me hours to develop and craft, and I had stumbled on to it entirely by chance.
As satisfying as this was there was there was a thin layer of anxiety that came with it. How did the artificial intelligence from Spotify figure out this list? How could it so effortlessly deliver on what I didn’t even know I wanted? Are we already at the place where Skynet knows more about me than I do about myself?
There’s one way to find out: use AI to help out on Tech Magazine.
You can see the results of this “collaboration” at five different places in this magazine. One of them is the editor’s letter on page 3, which replaced the editor’s letter you’re reading now. Try and find all five of the AI-generated pages in this issue and or scan the QR code on the right to find out how we did it.
Enjoy the AI-zine.
Tech Chief
We typed “Write 120 words about Winning Fingers Flashdash 4-In-1 Electronic Game” and the results are on page 40, with the human-written version next to it.
FLASHDASH 4-IN-1 ELECTRONIC GAME
Looking to pass the time without screens? FlashDash offers four quick-fingered games in a unique light-up grid design.
This durable plastic cube plays four different games using lights and sounds, for one person, two people or more. Great for home and travel, this fun hand-held toy helps improve brain skills and hand–eye coordination.
It includes a simple speed game called Chase Me, as well as popular classics like Simon Says. They’re easy to learn, but tough to master. This electronic game cube offers just the right amount of challenge for most people.
The perfect gizmo for road trips and family holidays, some would see this as a welcome break from screen entertainment.
R620 (import), winningfingers.com
DRAWING WITH AI
We also made artwork using AI. Below are two of the covers we created using the world’s leading imaging AI called DALL-E 2. In the end we didn’t use these covers, but the results were certainly usable, and were seriously considered.
We typed “need for speed unbound” into DALL-E 2 and it drew sports cars like this…
We typed “artificial intelligence” into DALL-E 2 and this was the result:
To understand how DALL-E works, read our story on page 15.
PLAYING AROUND WITH AI
Here are other things we asked ChatGPT about which were not used in the magazine, because it contained incorrect or misleading information. We typed:
1. “what are the specifications for an entry-level gaming PC?”
2. “Who is the greatest Springbok player of all time?” and
3. “is AI dangerous?”
Here are ChatGPT’s answers:
“what are the specifications for an entry-level gaming PC?”
[Answer is not bad, but too expensive to recommend for SA gamers]
An entry-level gaming PC typically consists of the following specifications:
CPU: An AMD Ryzen 3 or Intel Core i3 processor is sufficient for entry-level gaming. These processors offer good performance at a reasonable price.
GPU: A dedicated graphics card is essential for gaming. A budget-friendly graphics card such as Nvidia GTX 1650 or AMD Radeon RX 570 can provide good performance for entry-level gaming.
RAM: 8GB of RAM is typically enough for most games. However, some newer games may require more, so it is recommended to have 16GB of RAM for optimal performance.
Storage: A combination of solid-state drive (SSD) and hard disk drive (HDD) is recommended. An SSD can provide faster loading times for games and the operating system, while an HDD can offer more storage space at a lower cost.
Power supply: A 500W power supply is recommended to provide enough power to the components.
Motherboard: A budget-friendly motherboard that supports the processor and graphics card is sufficient for entry-level gaming.
Cooling: A budget-friendly CPU cooler and case fans can provide adequate cooling for an entry-level gaming PC.
These specifications can be adjusted depending on the user’s specific needs and budget. For example, a more powerful CPU or graphics card may be needed for more demanding games, while additional storage may be required for storing game files and other media.
——–
“Who is the greatest Springbok player of all time?”
[Notice how ChatGPT incorrectly states that South Africa won the Rugby World Cup in 1987.]
There have been many great Springbok rugby players throughout history, so it is difficult to identify just one as the “greatest” of all time. However, here are a few players who are widely regarded as some of the best Springboks ever:
Francois Pienaar: As captain of the Springboks, Pienaar led the team to victory in the 1995 Rugby World Cup. He was also known for his aggressive and tenacious play as a flanker.
Danie Gerber: A speedy and skilled center, Gerber is regarded by many as one of the greatest backs in Springbok history. He played in the 1980s and was known for his exceptional try-scoring ability.
Naas Botha: A legendary flyhalf, Botha was known for his kicking prowess and tactical acumen. He played for the Springboks throughout the 1980s and was a key part of the team that won the Rugby World Cup in 1987.
Bryan Habana: A winger with lightning speed and a powerful running game, Habana is one of the most prolific try-scorers in Springbok history. He played for the team from 2004 to 2016 and was a key part of the squad that won the Rugby World Cup in 2007.
These are just a few examples of the many outstanding players who have represented the Springboks over the years. Each player brings their own unique skills and contributions to the game, and their legacy lives on as an inspiration to future generations of rugby players.
——–
“is AI dangerous?”
Artificial intelligence (AI) can have both positive and negative impacts, depending on how it is developed and used. There is a growing concern among experts that certain AI systems may have the potential to cause harm, either intentionally or unintentionally. Here are some of the ways that AI could be considered dangerous:
Unintended Consequences: One of the main concerns with AI is the risk of unintended consequences. This can occur when AI systems are designed with unintended biases or errors, or when they behave in ways that were not anticipated by their designers. In some cases, these unintended consequences can have serious or even life-threatening consequences.
Misuse: Another concern is the potential misuse of AI by individuals or organizations with malicious intent. For example, AI systems could be used to spread misinformation or propaganda, to automate cyberattacks or other forms of digital warfare, or to create autonomous weapons that could cause harm without human oversight.
Job Displacement: AI could also be considered dangerous in the sense that it could displace human workers and lead to social and economic instability. If large numbers of jobs are automated, it could lead to mass unemployment and economic inequality.
Privacy and Security: Finally, there are concerns about the privacy and security implications of AI. As AI systems become more powerful and ubiquitous, they could be used to collect and analyze vast amounts of personal data, which could be used for nefarious purposes or could be vulnerable to hacking or other security breaches.
It’s important to note that many experts believe that the benefits of AI far outweigh the potential risks, and that the technology has the potential to solve some of the world’s most pressing problems. However, it’s also important to proceed with caution and to carefully consider the potential risks and unintended consequences of AI as it continues to develop and advance.
THINK YOU’VE GOT A GRIP ON AI? HERE’S A BONUS GEEK TIP TO SHARE AROUND THE BRAAI:
