Most work is new work!
Apr 16, 2024 12:30 pm
Hey,
it's been a while! And it's not you, it's me! I had a book to publish (Dutch translation of Equal Pay for Equal Work), a couple of keynotes to prepare, and as you receive this I am en route to FAIR Barcelona. In two weeks I am a panelist at HR Tech Europe and then it's off to Unleash in Vegas, where I'll host a round table on technology for Pay Transparency (sold out!). I'll also present a session on payroll and integration. Come say "Hi" when you're there!
I always collect links to articles I want to read when I have time. Or podcasts I want to listen to. This time I'm going to share 5 links that captured my attention. I'll start with 3 on AI at work, followed by 2 on the future of payroll, just to balance it out. Let me know which one is your favorite!
1 Most work is new work!
I have a slide with news paper headlines starting back in 1950 announcing the end of work. Since then, we have more, not less work.
The current hype around generative AI claims that we might not have jobs in a few years. Or maybe I should say: that 'work' won't exist. And then I watch another AI experiment go off the rails, like the NYC chat bot, and I am reassured we'll need experts for the foreseeable future...I'm not too worried about the future of my work, and I don't think you should be. Will work be different? Undoubtedly. Just like it has changed for decades. Will it disappear? Highly unlikely.
I've also noticed that when tech bros claim AI will replace us all, everyone nods their head. No one asks them: but how? No one ever asks about the details. Do they have the slightest idea what people actually do for a living? I'm not sure. But after watching my plumber work, I am even more curious how their work will be replaced by AI before 2030. If you have the answer, let me know!
MIT News took a look at new research from Ford Professor of Economics David Autor that is focused on new job creation in the United States since 1940:
“We estimate that about six out of ten jobs people are doing at present didn’t exist in 1940,” says Autor. “A lot of the things that we do today, no one was doing at that point. Most contemporary jobs require expertise that didn’t exist back then, and was not relevant at that time.”
A lot of that has to do with new technology, and the need to adjust roles as a consequence. Or create new ones. But it's not just technology that creates new roles, it's also demand: for services, like a conference planner or a tattoo artist. It's a fascinating article that also looks at the potential effect of AI on work. Read Most work is new work: Part 1 and Part 2.
2 Could AI rebuild the middle class?
The same David Autor also wrote a thought-provoking article titled AI Could Actually Help Rebuild The Middle Class. If you’ve attended one of my keynotes, you’ve heard me speak about demographics. And you've seen a population pyramid that shows more people leave the workforce than join it. The labor shortage is real and here to stay for much longer than we assume.
That’s why this article resounded so much with me. He writes: “The industrialized world appears poised to run out of workers before it runs out of jobs.” And so, advances in new tech like AI couldn’t have come at a better time. He describes how AI can help inexperienced workers get up to speed quickly, why AI could actually lead to better jobs - helping people to become experts - and help rebuild the middle class. He uses some recent research to underpin his assumptions.
I'm left wondering though how companies would respond: if everyone is an expert, wouldn't that be a reason to compress all salaries into an expert level? And would the labor shortage really compel companies to pay workers more? As we've seen over the past decades, it's typically not the workers who are the beneficiaries of higher margins and profits.
If you have some time, read the article. It’s more upbeat than anything I’ve read so far about AI and the workforce. And it will expand your thinking about the most productive use cases of AI in the workplace.
3 What if we run out of LLM training data?
I did not realize this was a thing, but apparently it is. Because Large Language Models have gobbled up almost all available data on the internet, they are running out of it. And that's a problem. You must constantly train LLMs to improve them, and you need fresh data for that. When you run out of data, where do you turn?
Maybe you could use the output: LLM-generated data, so-called synthetic data. But that would not necessarily make an LLM smarter. In fact, researchers found that feeding synthetic content to AI models causes their output quality to erode: "without enough fresh real data [...] future generative models are doomed to have their quality (precision) or diversity (recall) progressively decrease". The researchers noted this will drive the model MAD.
I learned about this phenomenon while listening to the Hard Fork podcast on AI's Data Wall, a fascinating topic. You should also translate it to your own situation: if you are building an in-house LLM, do you have enough data to train it? What happens when the data runs out? Will your LLM become dumber, not smarter? And then what? Would you need a human to replace AI because it's not so intelligent after all?
4 Is global payroll truly global?
Companies say they have global payroll, but when you look under the hood, it's often a collection of local payrolls stitched together via an integration layer. That has benefits, like centralized control and global reporting, but it also comes with risks, like security, compliance and costs because of the # of parties involved. It can also negatively impact the employee experience. The challenge of managing this setup is a major frustration of payroll leaders.
Deel commissioned Forrester to do a deep-dive into the benefits of a global payroll solution: Is Global Payroll Truly Global. Forrester interviewed 300+ payroll leaders and some of their findings are surprising. Companies still don't have a good view on the costs of their international payrolls. 85% of respondents think that payroll costs will rise. It means they don't quite have a handle on total costs, which is worrisome. Especially in light of my observation that compliance is only getting more complex.
In addition, 63% of HR and finance leaders aim to improve their global payroll efficiency and workflows. Even after all the optimization projects of the past years, there is still room to improve.
Shortly after I received the report, Deel announced that they will acquire Payspace. A young EoR vendor buying an incumbent! I think they have a great opportunity to shake up the payroll industry. It inspired me to write an article about the state of the payroll market, and why we need more change.
5 What is the Future of Payroll?
I participated in HR.com's advisory board for a new report called Future of Payroll: New Trends 2024. With so much change from evolving regulations, new technologies and evolving payroll offerings, how can payroll leaders make sense of all of this? The study aims to help payroll leaders understand what their peers are experiencing and what they view as most important. This can help you determine your priorities for the next years.
Based on HR Research Institute’s Payroll Maturity Model, we asked the respondents to rate themselves: only 16% of the responding organizations rated themselves as highly strategic. I am curious to learn how you would rate your payroll service? Have you achieved a highly strategic way of working or is there still too much administration to be done? You can download the report or watch the webinar where I was interviewed about the findings.
Have a great day, Anita
How to get in touch
My newsletter is free: please think of me when you need an engaging and insightful keynote or workshop!
Or support me by buying my books Equal Pay for Equal Work and How to Select Your Next Payroll. Or get the Dutch translation Gelijk werk? Gelijk loon!
(Don't forget to switch to your local Amazon store)
If you need HR Tech or payroll advice, please connect with me here:
What can I do to make the newsletter better for you?
Did you enjoy this newsletter?
Support me and forward it to someone who might like it.