Discover more from FullStack HR
The AI revolution your HR department has been waiting for.
I'm pretty sure we will be partly replaced by bots.
Tired of bland HR advice and the same old, same old? FullStack HR always tries to be forward-leaning and takes a solid stance in HR.
Join the over 2000+ like-minded HR professionals who subscribe to FullStack HR by subscribing below.
Long-time and no see. As per usual in December, life got in the way.
Or, to be more precise, life, Midjourney, and ChatGPT got in the way, and I'll come back to the later in a minute or two.
It feels good to be back publishing again, and I look forward to 2023 from a FullStackHR perspective.
Closing out 2022, I'm also thankful from the bottom of my heart for all of you who take time every week to read this newsletter. It means a lot to me, and I am eternally grateful. So thank you!
During the holidays, I’ve tried to set the direction for the year ahead by looking at the year that passed.
The most popular articles from 2022 were:
It looks like you all enjoy reading how-to articles.
My goal for 2023 is to continue publishing every week and to be an optimist in the future of work and HR. I may not always get things right, but I hope to provide thought-provoking content and encourage critical thinking.
The comment section at the bottom is always open for you to ask questions, share your views, and provide feedback to the community and me!
Let's move on to this week's topic and the first article of the year: ChatGPT, what it is, and potential use cases for HR.
It will be a series, and next week, we'll focus on practically using ChatGPT for HR tasks.
If you want to learn more about AI and HR, I’m planning an HR AI Roundtable group where we’ll meet and discuss the future of AI and HR.
Drop your email here if you are interested in joining!
Ok, ready? Ready!
In early December last year, the internet was excited in a way that hadn't been seen in almost a decade.
What was the cause of all this hype?
ChatGPT by OpenAI.
ChatGPT is an online language generation system that uses deep learning to generate text in an open-ended style similar to a conversation. To better understand ChatGPT, let's start with some background information about OpenAI and the organization behind it.
OpenAI is a research organization focused on developing and promoting friendly artificial intelligence. It was founded in 2015 by tech leaders and researchers, including Elon Musk and Sam Altman, to advance AI in a way that benefits humanity. OpenAI researches various areas, including machine learning, robotics, and economics. It has developed several influential AI technologies, such as GPT-3, a state-of-the-art language processing system.
GPT-3, you say? But what is GPT, and what does the "3" stand for?
GPT-3 stands for "Generative Pre-trained Transformer version 3." It is a state-of-the-art language processing system developed by OpenAI and is one of the largest and most powerful language models with a capacity of 175 billion parameters.
GPT-3 can be used for various natural language processing tasks, including translation, summarization, question answering, and text generation.
What makes it so powerful is that it was trained using a technique called reinforcement learning, which involves using trial and error to train a model by observing the consequences of its actions. In other words, the system learns through interaction with its environment and doesn't rely on human input or programming, making it flexible and versatile. This makes it well-suited for use in chatbot applications and other applications that require strong language modeling.
Going back even further, GPT-3 builds on a paper by Google outlining the Transformer model. The Transformer model is an artificial neural network that processes sequential data such as natural language text.
It uses a self-attention technique to weigh the importance of different elements in the input sequence, allowing it to process the sequence more efficiently and effectively. This makes it especially useful for longer sequences and has been successful in several natural language processing tasks. The Transformer architecture also includes innovations like multi-headed attention and position-wise feed-forward layers, which help the model understand the input data better.
Put in picture language, the Transformer is a machine that can build structures using blocks of different shapes and colors. It uses self-attention and other techniques to choose and arrange the blocks, allowing it to quickly and easily create many different structures. GPT-3 is a super advanced Transformer with a large capacity and the ability to learn independently. This is useful for chatbots and other tasks that require strong language skills. (And yes, I used ChatGPT to generate this!)
I say this because I think it's important to use and experiment with an application like ChatGPT to understand how the model works. This is a very simplified version, but it illustrates the core concept and what’s behind the names and models.
But enough about the technical details - let's talk about why people are so impressed by ChatGPT.
No, not another chatbot.
The idea of chatbots has been around almost as long as the internet itself. Over the years, many attempts have been made to bring this futuristic vision to life, but all have failed.
Most of us have probably experienced talking to a "customer support" chatbot and been extremely frustrated by the generic responses that don't provide any real help.
ChatGPT is nothing like these chatbots.
It's the most human-like experience I've ever come across. You type something in, and you get a thoughtful response back. It also remembers what you wrote previously, which makes a huge difference and creates a stronger connection between you and the machine. It’s like chatting with someone.
People have been using ChatGPT in various ways - it has helped engineers with coding problems, acted as a psychologist, created content calendars, and even written essays. When I played around with ChatGPT, I was able to create everything from a 20-week marathon training schedule to a short story about a sea monster living beneath the Atlantic Ocean. The possibilities seem and feel endless, hence the hype.
But enough about sea monsters - this is, after all, a newsletter loosely connected to HR.
How will ChatGPT impact HR?
There's been a lot of debate about whether AI will eventually make human work obsolete. While I believe that humans will always be needed, AI will significantly increase our productivity and reduce the need for headcount in the future. ChatGPT shows a glimpse of that, a glimpse of the future.
Fewer people will be able to do more.
This won't happen overnight, but if you consider the rapid pace of technological development, it's likely to happen in the not-so-distant future.
Unlike what most gurus previously believed, white-collar jobs will probably be impacted first.
Why call a labor law lawyer when you can talk to an AI-generated one that instantly provides the help you need and is up to date with the latest law changes and interpretations from all courts around the world?
Or why hire an HR Business Partner to support your business when you can get the same help from an AI HR Business Partner that knows all your documents, policies, and laws in all the countries you operate in and can answer the manager's questions in their local language?
Based on employee surveys and insights from looking at who's emailing who and talking to whom on Slack, the AI HR Business Partner can suggest what the manager can do to improve team communication.
Why would you then need an army of HR Business Partners?
You'll still need a few to handle corner cases and adequately train the model, but we can be sure that we will need fewer in the future.
And I can already hear someone screaming from the back about “the strategic impact” and that “AI will never be able to take all aspects into account.”
As per above, if we grant the HR-AI-assistant-Business-Partner access to our digital work lives, they will have more information and data available than we ever could dream of as HRBPs. Did I also mention that Microsoft is heavily involved in OpenAI? No? They are and will most likely use the technology in their products.
We also need to consider that if we need fewer people in our organization, there will also be less need for all aspects of HR.
Am I worried about this? Not really. Throughout history, humans have found new things to do and roles to fill. We will do this again, but if you think your job today will look the same ten years from now, you need to think again.
And does all of this sound too futuristic? For me, it certainly did up until the 30th of November.
Then things changed, and I have to admit that I underestimated how quickly we were moving towards this future, even though books like Life 3.0, Abundance and The Second Machine Age had helped paint a picture of a future society where an increasing number of people will be working less. I still underestimated the power of the shift.
How can you use it?
Next week’s article will be all about this, with examples and prompts to use. So view this as a sneak peek.
First and foremost, it's super easy to start using ChatGPT, and it’s free to use. As they are growing at an insane pace, their backend might struggle to keep up with demand from time to time. Don’t give up; try again.
Once you’ve signed up, you can start asking questions as you would do to a friend or colleague.
Remember to avoid sharing sensitive information, as everything is stored and analyzed on OpenAI's servers to train the model further.
Here are some examples of how you can use ChatGPT if you work in HR:
Providing information and answering questions: ChatGPT can provide general information and answer questions on a wide range of HR-related topics, such as employment laws, best practices, and industry trends.
Generating content: Create HR-related documents and content, such as job ads, employee handbooks, training materials, and policy documents.
Analyzing data: Analyze data on factors such as employee performance, turnover, and engagement to identify trends and make recommendations for improvement.
Assisting with communication: Draft emails, presentations, and other communication materials related to HR.
Remember, you shouldn't share sensitive information, but as you can see, your imagination is the only thing holding you back, not the tool itself.
Should you believe the hype?
Typically I usually advise people not to jump on hypes. They are seldom beneficial for the average person or organization. But this is different.
This is a hype you definitely should jump on and start exploring today. Now.
AI tools like this will impact us all and happen fast. Start learning now.
Read books. Talk to people. Try the tools.
I’m planning an HR AI Roundtable group where we’ll meet and discuss the future of AI and HR. Drop your email here if you are interested in joining!
Next week I’ll show more practical examples of how you can use the tool - stay tuned!