By Thalif Deen
UNITED NATIONS, 16 May 2023 (IDN) — The world is currently experiencing one of the fastest advances in modern technology: the AI Chat GPT search engine, which has generated both curiosity and skepticism.
At a White House May 4 meeting of executives from Google, Microsoft, Anthropic and OpenAI, the maker of ChatGPT, US President Joe Biden conveyed mixed feelings: “What your’re doing has enormous potential—and enormous danger”.
There is both an upside and a downside to ChatGPT while a raging debate continues on both sides
Currently Chat GPT is being deployed to crank out email messages, articles, pitches and reports with little or no human inputs.
In a report last month, Cable News Network (CNN) said “a chatbot powered by artificial intelligence, can pick stocks better than your fund manager, according to analysts”.
A recent experiment found that the bot far outperformed some popular UK investment funds—and funds managed by HSBC and Fidelity were among those selected.
Between March and April, a dummy portfolio of 38 stocks gained 4.9% while 10 leading investment funds clocked an average loss of 0.8%, the results showed.
The analysts asked ChatGPT to select stocks based on some common criteria, including picking companies with a low level of debt and a track record of growth.
Microsoft, Netflix and Walmart were among the companies selected. While major funds have used AI for years to support their investment decisions, analysts say ChatGPT has put the technology in the hands of the general public—and it’s showing it can possibly disrupt the finance industry.
The rapid advances in technology have also impacted on the recent strike by the Writers Guild of America which virtually brought late night TV shows to a complete standstill.
A TV and film writer was quoted as saying “part of what led to this strike is that studios are acting like tech companies—trying to optimize by replacing humans with computers.”
In an interview with IDN, Eric Grigson*, Director of Developer Experience at the Melbourne-based Culture Amp, said “the first thing that is important to understand when we are discussing chatbots like ChatGPT is how they work”. https://www.cultureamp.com/
ChatGPT is a program, he pointed out, that is designed to take a prompt given by a user and respond with strings of words that it predicts will best satisfy the user, based on the words that it was trained on.
“It isn’t actually capable of thinking or understanding language—everything that it does is based on probability and numerical analysis. In other words, ChatGPT is a program that guesses what you want it to say, based on what you ask it. It is very good at this, but that doesn’t mean it gets everything right”.
This is partially because it cannot think, Grigson said, but also because of the data it was trained on (random junk from every corner of the internet). ChatGPT is subject to all of the inherent biases that live on the internet, and the sheer volume of data it is trained on (ChatGPT uses more than 500 billion data points—referred to as tokens) makes it impossible to avoid this. This is a big part of why a lot of folks consider current generation AI “dangerous”.
It is sophisticated enough, he said, to appear as if it is capable of thinking, but it really isn’t. Instead, when you ask a question, it crunches the numbers and gives you a response which seems like it might be right, but could also include some hefty inherent biases. ChatGPT is a really smart word generating machine—it is not intelligent.
“I use it myself whenever I need to write something for a large audience. I write a rough draft in my own words, and then feed it into ChatGPT and ask it to rewrite the text in a more concise or clear manner. I then clean up what it has written (it never gets things perfect) and then share it out”.
“It’s a huge time saver, as I spent less time thinking about how to phrase things clearly, and more time thinking about what I want my audience to think and feel when they are reading my words”.
“I think helping folks quickly write content like this is the big impact it can have. There are a lot of clever applications of ChatGPT that can help people write faster and save them time, and in this way, I think it will absolutely impact the world of work.”
Excerpts from the Q&A:
Q: What about its impact on a world body like the United Nations—and whether, in the long run, it could result in the replacement of humans (read UN staffers)?
A: “AI chatbots cannot think, they can only spit out words based on prompts from users. This makes it very unlikely that AI would wholly replace UN staffers—as staffers can think!”
“However, it is likely that as AI chatbots make it easier to produce quality written content, the general standard of communication in the workplace will rise. In other words, folks won’t get replaced, but we will likely see the general standard of communication grow, and some people may get left behind if they cannot keep up.
Q: What jobs can be taken over by AI?
A: AI is good at producing words quickly based on what it is trained on. You can train an AI how to understand how something works (even something very complex, like a computer), and then it will be able to answer questions and explain things to people at length in an interactive personal context. A good application of this might be in customer support, where an AI could replace front line support workers who deal with basic questions.
I don’t see AI replacing most roles that you might find in the UN though, as it isn’t capable of thinking, only regurgitating data. It cannot think critically or creatively, and it cannot innovate. It can fulfil some basic requirements, but it cannot inspire, thus it’s impact will be less than a human’s.
Q: Will most future reports and publications be AI-assisted? Is this also possible with the annual reports of the UN Secretary-General?
A: We are going to see a lot of reports and publications come out in the next few years that were crafted with the help of AI tools, and we’ll be better off for it. People can think more about what they want to say, and AI can help them find the right words to efficiently express themselves and communicate.
To your question, I think it is very likely that the annual report of the SG could be produced with the help of AI in the future (provided the UN is comfortable with tech companies having access to the entire contents of the general report before it is even published!).
In the long run, tools like ChatGPT will not replace us. They will make us more efficient when it comes to writing and communicating. As they evolve, they will become as commonplace as mobile phones or keyboards. Everyday tools that we think nothing of.
However, I do think that AI chatbots have a fundamental impact on jobs in the long run. As written communication becomes more efficient and AI takes that burden off of our shoulders, I think the value of creativity and having a unique perspective will become more and more important.
AI will effectively raise the bar, and the standard required to hit success and achieve high performance will grow higher. For some, this will be a fantastic opportunity to grow and evolve. Others will get left behind, unless they learn to adapt and outthink AI.
* As the Director of Developer Experience, Eric Grigson is experienced in leading teams of engineers in delivering complex technical projects. Additionally, his expertise in coaching and training has allowed him to design and deliver impactful courses for fellow employees, helping them learn the skills necessary to manage teams effectively. [IDN-InDepthNews]
Image by Alexandra Koch from Pixabay
IDN is the flagship agency of the Non-profit International Press Syndicate.
We believe in the free flow of information. Republish our articles for free, online or in print, under Creative Commons Attribution 4.0 International, except for republished articles with permission.