I’ve said for years that AI won’t steal your job, but that smart people who put AI to work will. But is that still true? With the emergence of DALL-E, Stable Diffusion, and especially, ChatGPT, aren’t at least some jobs at risk? There’s evidence that ChatGPT can get its MBA, do well on medical exams, and pass the bar. Surely, that means it’s ready to start taking all our jobs away, doesn’t it?
Well, maybe. But I’m still not convinced. Smart people who use AI — and, in particular, tools like ChatGPT — remain a much bigger threat than AI or ChatGPT alone.
What makes me think that? Why am not too worried about ChatGPT stealing my job — or yours? And, exactly how confident am I about that? That’s what this week’s episode of Thinks Out Loud is all about.
Want to learn more? Then give a listen, review the transcript, and check out all today’s show notes. Enjoy!
Is ChatGPT Going to Steal Your Job? — Headlines and Show Notes
Show Notes and Links
- AI Won’t Steal Your Job: Smart People Who Put AI to Work Will (Thinks Out Loud Episode 208)
- Will ChatGPT Kill Google? (Thinks Out Loud Episode 367)
- Is Artificial Intelligence Still a Big Trend for Your Business? (Thinks Out Loud Episode 311)
- The Most Important Trends in Marketing 2023 (Thinks Out Loud Episode 369)
- The Future Digital Trends You Must Think About Today (Thinks Out Loud Episode 342)
- Ghost Writer: Microsoft Looks to Add OpenAI’s Chatbot Technology to Word, Email — The Information
- ChatGPT is ‘not particularly innovative,’ and ‘nothing revolutionary’, says Meta’s chief AI scientist | ZDNET
- Would Chat GPT Get a Wharton MBA? New White Paper By Christian Terwiesch – Mack Institute for Innovation Management
- ChatGPT passes exams from law and business schools | CNN Business
- OpenAI has hired an army of contractors to make basic coding obsolete | Semafor
- Researchers Tested ChatGPT on the Same Test Questions As Aspiring Doctors
- ChatGPT passed an MBA exam and one professor is sounding the alarm | Fortune
- Inherent Trade-Offs in Algorithmic Fairness – Microsoft Research
- The secret bias hidden in mortgage-approval algorithms | AP News
- On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? PDF link
- Understanding the Capabilities, Limitations, and Societal Impact of Large Language Models PDF link
- CNET’s AI Journalist Appears to Have Committed Extensive Plagiarism
- LaMDA: our breakthrough conversation technology
- Google Calls In Larry Page and Sergey Brin to Tackle ChatGPT and A.I. Chatbots – The New York Times
- Google and Meta moved cautiously on AI. Then came OpenAI’s ChatGPT. – The Washington Post
- Gmail’s creator says ChatGPT will destroy Google in two years
- Revisiting How to Drive Business Using Digital in 2023 (Thinks Out Loud)
- Is It Time for Digital Marketers to Move On to The Next Big Thing? (Thinks Out Loud Episode 360)
- The End of Google? (Thinks Out Loud Episode 372)
- What We Learned from Big Tech’s Earnings Q1 2023 (Thinks Out Loud Episode 373)
- Big Trends: The Early Innings of AI in Marketing (Thinks Out Loud Episode 374)
- Big Trends: Does Marketing Have a Future? (Thinks Out Loud Episode 375)
We have some free downloads for you to help you navigate the current situation, which you can find right here:
- A Modern Content Marketing Checklist. Want to ensure that each piece of content works for your business? Download our latest checklist to help put your content marketing to work for you.
- A Brief Introduction to Thinks Out Loud. As a bonus, we’ve also included this PDF document that highlights some of our core episodes to help you dig into what the show is about. We think it will help you capture the show’s essence while you’re working your way through the 300-plus episodes published so far. Download it here.
- Digital & E-commerce Maturity Matrix. As a bonus, here’s a PDF that can help you assess your company’s digital maturity. You can use this to better understand where your company excels and where its opportunities lie. And, of course, we’re here to help if you need it. The Digital & E-commerce Maturity Matrix rates your company’s effectiveness — Ad Hoc, Aware, Striving, Driving — in 6 key areas in digital today, including:
- Customer Focus
Subscribe to Thinks Out Loud
Contact information for the podcast: firstname.lastname@example.org
Past Insights from Tim Peter Thinks
Technical Details for Thinks Out Loud
Recorded using a Heil PR-40 Dynamic Studio Recording Mic and a Focusrite Scarlett 4i4 (3rd Gen) USB Audio Interface into Logic Pro X for the Mac.
Running time: 22m 21s
You can subscribe to Thinks Out Loud in iTunes, the Google Play Store, via our dedicated podcast RSS feed (or sign up for our free newsletter). You can also download/listen to the podcast here on Thinks using the player at the top of this page.
Transcript: Is ChatGPT Going to Steal Your Job? (Thinks Out Loud Episode 371)
Well, hello again everyone, and welcome back to Thinks Out Loud, your source for all the digital expertise your business needs. My name is Tim Peter. This is episode 371 of the Big Show. And thank you so much for tuning in. I very, very much appreciate it. I know we missed an episode last week. There’s no real good reason for it other than “circumstances intervened,” but I do appreciate you sticking around and hanging with us in the interim. And I promise you we’ll have another episode for you next week. In the meantime, I think we’ve got a really cool show for you this week.
Is it Still True That ChatGPT and Other AI Won’t Steal Your Job?
A few years ago. I posted an episode that said AI won’t steal your job, but smart people who use AI will. And I don’t know if you’ve noticed, but with the emergence of ChatGPT and other Generative AI tools like Stable Diffusion and DALL-E and Midjourney, and the like, people are starting to freak out. Like there’s some really, really crazy stuff out there.
We’ve seen examples where ChatGPT has done very, very well on real world assessments of human competence. So for instance one Wharton professor showed that ChatGPT could attain an MBA. We’ve seen examples where you know it could potentially attain a medical degree. We’ve seen examples where it could pass the bar. There literally was a story just the other day where someone was going to use an AI as their lawyer till a local bar association stepped in and explained why that’s a really, really bad idea and that they might be committing legal fraud. But results like these worry educators.
Given that, it’s reasonable to ask today, again, is it still true that AI won’t steal your job? And is it only going to be “smart people who use AI” or is AI itself coming for your job? Should students who are thinking of going to business school or law school or medical school seek another line of work? What about the educators who teach those folks? Is there going to be a role for marketers or digital strategists or, oh, I don’t know, consultants and authors who focus on those groups when we have machines that might be able to do those jobs equally well?
How Confident Should We Be That ChatGPT Won’t Steal Your Job?
And the first thing I want to say is — and, you know, this show is called Thinks Out Loud for a reason — my answer at the moment is, “I’m not sure.” I’m not a hundred percent sure.
I’ve talked about this before, that when you talk about where we’re going to be in the future, first you should define what you mean by the future. How soon do you define that timing? You should talk about, and you should think about what do we mean by success? Or in this case, “steal a job.” What would that actually look like? And we should also really get a good feel for what is our confidence level? How much would we be willing to bet on a particular answer?
And so everything that I’m going to tell you right now, my confidence level is moderate. I want to start there. I’m not a hundred percent sure. I’d bet decent money that I’m right, but I wouldn’t bet you know, huge money because this is changing day by day. ChatGPT, for instance, is built on a platform called GPT-3.5 and GPT-4 is coming somewhere not too far down the road and could upend everything that we know right now.
At the same time, I’ve had the privilege over the last several years of working with a couple of startups focused on applying artificial intelligence and machine learning and natural language processing to solve various marketing and customer service challenges.
AI Has Been Around for a While… Shouldn’t It Already Have Won?
Many of the techniques themselves have been around for a while. Facebook’s head of AI just said the other day that what they’re doing with ChatGPT is not particularly “computationally innovative.” Now he’s taking some real heat for that because I mean, technically, by that reasoning, the iPhone wasn’t terribly innovative either; it just put together a lot of things that already existed. However, they also did it in such a way that it completely changed the way phones work. So you don’t always have to be the coolest new tech to win in the marketplace. You have to bring it to market in a way that works for customers. Execution matters. So, you know, we may be seeing tech that’s good enough and just needs to be executed better before this becomes something that we need to worry about.
The other thing I would note is Microsoft is making huge investments here. It has long been an investor in OpenAI, which is the company, the organization, behind ChatGPT, GPT-3.5, DALL-E. So some of these tools, Microsoft has been really, really invested in. Quite literally. Now they’re putting in I think $10 billion. So not a trivial sum. And they expect to incorporate these tools into their Office suite and into their search results on Bing. They think it’s going to be a big deal.
What Microsoft’s Investment in OpenAI Suggests
But the way they’re investing is one of the things that gives me a little bit of hope that you probably aren’t going to lose your job to an AI. They intend to build these tools into Word and into Outlook and into PowerPoint and into Excel. They’re not saying Word is going to write the whole document for you, or Outlook is going to write the whole email for you, or PowerPoint is going to create the entire slide, the deck for you. They’re saying these are tools that you are going to use to create your Word documents, to create your emails, to create your slide decks. So I don’t think I don’t think Microsoft — who should know a lot about this — thinks that’s going to actually come and completely, completely upend the world.
Where Will Training Data Come From?
Another reason that I have some hope here and why you should have some hope is that these types of tools are trained on content. They’re trained on really high quality, expert level data. That’s how they get good at what they do. Well, if they’re producing data that is frequently confidently wrong, you can’t use the output of the system to improve the system. You won’t make the system smarter. You’ll make the system dumber.
They’re still going to need high quality, rich, expert content to train these systems. If you teach the machine less skillfully, less capably, their outputs will get worse. And that means that people who have expertise, people who have knowledge — and can communicate that knowledge — still will have tons and tons and tons of value, probably for a long time to come.
And I’m going to come back to that in just a moment.
The AI Bias Problem
One more challenge of course, is that because of the way these systems are programmed, trained, because of the way they’re developed, they frequently and easily incorporate bias. You’ve probably heard the story about Microsoft exposing an AI powered bot called Tay to Twitter, only to have Twitter turn sweet, innocent little Tay into a ranting, racist, sexist homophobe in about 16 hours. More recently, Facebook thought that they’d solved for that problem. And like Facebook does, they didn’t really learn from Microsoft’s mistake. They put their system out there — Blender Bot 3 — out on the internet. And the internet did fail to make the bot anti-Semitic… for a whole weekend. That quickly turned into a disaster much like Microsoft had discovered with Tay. So, You know, this is something that is a real problem for these systems, that they incorporate bias incredbily easily.
One of the reasons the researchers who created GPT-3 have said that the technology “exhibits several racial, gender, and religious biases,” at least in part because those biases are baked into the content used to train the models. In one example that I’ve seen they talked about how the machines kept finding terms like “women doctors,” and that caused the systems to “learn” that doctors aren’t women. Which of course is not true. But because of the way they’re trained, a lot of these biases are baked in. We’ve seen examples where tools used to approve mortgages were much more likely to reject loans from Latino, Asian/Pacific Islanders, native Americans, and Black applicants because of bias that they picked up. We’ve seen situations where sentencing tools in court systems gave longer sentences to people of color than they did to people who are white for instance.
Will AI Always Exhibit Bias?
Now there’s a wild reason why that is, and it might be that some of these biases may be impossible for AI to solve. Not difficult. Impossible. Microsoft Research has talked about how there are “inherent trade-offs” in algorithmic fairness. There was an interesting talk that’s out on YouTube. You can watch the whole thing. It’s interesting. It’s pretty heavy. And to be fair, I understood about one word in three, but the basic premise was that it is very, very challenging to develop systems that are inherently fair.
I’ll give you an example of that. Define “fair.” Define “just.” Define “legal.” There are so many possible definitions to these. We have many arguments as people about what is fair, what is just, what is legal — constantly. Otherwise we wouldn’t need negotiators or mediators or lawyers or judges or priests or rabbis or people to help us understand what these things are.
It is unlikely to realistically expect machines to agree on what is right when all of humanity struggles with those concepts pretty much every day. That’s not a computer science problem. That is a human behavior problem. And don’t get me wrong, obviously researchers are doing things to try to correct for these biases, but they may just be baked into the way that people work, let alone machines.
What Google’s Experience Suggests
And as long as the inputs derive from human human knowledge and human experience, then it’s pretty likely AI will continue to echo those biases. Google has said that they have similar technologies as ChatGPT, such as LaMDA, but they haven’t wanted to expose them to the world because of brand risk. Like they’re worried people are going to see these and it’s going to make Google look bad.
And yeah, yeah, yeah; I get it. They make their money from advertising and it’s entirely possible that they haven’t figured out a way to monetize these systems either. But given the experiences we’ve seen from people like Microsoft with Tay and like Facebook with Blender Bot 3, I’m willing to give them at least a little bit of doubt on this one, that this may be a tougher problem to solve.
ChatGPT and AI Might Not Steal Your Job, But They Might Take Away Key Tasks
Now, I don’t want to give you the wrong impression here. You know, it’s easy to listen to what I’ve just said and go, “Oh, cool! AI can’t possibly take my job!” Well, that’s not true either. You know, every silver lining has a cloud, right?
The thing you want to, have to, realize is AI is baked into many technology platforms around marketing, around pricing, around revenue optimization, around customer relationship management, around workflow optimization, around cybersecurity and analytics. These things aren’t “coming.” They’re here. Now. And they’re getting better every single day. So I don’t think you should look at this and say, “Oh, they’re no threat at all.” They are.
I would also encourage you to think instead about three questions.
- One of these is what tasks, not jobs, but tasks, can AI do better than people? Where does AI have an advantage today?
- I want you to think about what tasks can people do better than AI, and where do human beings have an advantage?
- And most importantly, what tasks can people and AI working together do better than either alone?
Using AI to Automate Tasks Might Take Your Job Away
Because I think what is incredibly likely — the thing that’s going to challenge people the most — is not that your job’s going to go away. But, as stated previously, that people who use AI to farm out particular tasks where it’s really strong are going to be able to compete more effectively than you. They’re going to be able to do what they do faster, better, cheaper, more repeatably. And over time, those tasks may either be outsourced to AI — or the people who do those tasks using AI better than you — are going to be better at their jobs. They’re going to be better at finding and acquiring customers. They’re going to be better at serving customer needs. They’re going to be better at helping customers achieve what they want. And they’re going to steal business from you.
AI, Smart People, All That Jazz
So when I think back on the podcast episode that stated unequivocally that AI won’t take your job, but smart people who use AI will, I think that’s still very true. I think that’s still the world in which we live, and that’s the world in which we work, and that’s the world in which you have to continue to learn. So my advice, my suggestion, my thoughts on this are that that’s where you need to be focused. How do you use AI to do a better job at the things that you’re already trying to do, the things that AI does well? And where it does provide you a competitive advantage consistently?
AI will make some jobs go away. We’ve seen that lots and lots and lots of other times in the past. You know, my grandmother was a telephone operator. I don’t know if you’ve ever met any telephone operators in your life, but in my life I’ve met very, very few because they don’t really exist as a job any longer. Automated telephone switching systems made that job unnecessary.
AI can make some jobs go away. That’s definitely true.
Conclusion: Is ChatGPT Going to Steal Your Job?
But the bigger threat, the bigger risk, the bigger challenge to you is not that AI will take your job. It’s that people who know how to apply AI to solving real business problems and real customer problems are the ones who will take your job. The technology is not going to go looking for a job. People are going to go looking for jobs that AI can do better than you can do alone. So I think we’re still very much in a place where AI won’t take your job, won’t steal your job, but smart people who use AI will.
And the question you should be asking yourself is, “How do I become one of those smart people who use AI?: Because the job you may end up taking may be your own. But by thinking about how to use AI correctly, the job you may save is also your own. That might not be the job you have right now, but it could be the job you have for the rest of your career.
Show Closing and Credits
Now looking at the clock on the wall, we are out of time for this week. I want to remind you that you can find the show notes for today’s episode, as well as an archive of all past episodes by going to TimPeter.com/podcasts. Again, that’s TimPeter.com/podcasts. Just look for episode 371.
Subscribe to Thinks Out Loud
Don’t forget that you can click on the subscribe link in any of the episodes you find there to have Thinks Out Loud, delivered to your favorite podcatcher every single week.
You can also find Thinks Out Loud on Apple Podcasts, Google Podcasts, Stitcher Radio, Spotify, Overcast, anywhere fine podcasts are found.
Leave a Review for Thinks Out Loud
While you’re there, I would also very much appreciate it if you could provide a positive rating or review for the show. Those ratings and reviews help new listeners find the podcast. Those ratings and reviews help new listeners understand what the show is all about, and they help get the word out. They help grow our community and it means so very, very much to me that you helped do. I really appreciate you helping to make Thinks Out Loud a better place for everyone involved, so thank you so very, very much for that.
Thinks Out Loud on Social Media
You can also find Thinks Out Loud on LinkedIn by going to linkedin.com/TimPeterAssociates. You can find me on Twitter using the Twitter handle @tcpeter. And as always, you can email me at email@example.com. Again, that’s firstname.lastname@example.org.
Finally, I just want to say one more time how very much I appreciate you tuning into the show every single week.
I would not do this show without you. I don’t do this to talk to myself. I do this to have a conversation with you and build this community together. And it just means the world to me when you listen and you comment and you send me emails and that we get to have a dialogue about it. So please keep the emails coming. Please keep the discussion going. I do look forward to it.
With all that said, I hope you have a great rest of the week. I hope you have a wonderful weekend ahead, and I will look forward to speaking with you here on Thinks Out Loud next time. Until then, please be well, be safe, and as always, take care, everybody.