I shared an opinion last week on LinkedIn (original post here) about the potential impact of AI on knowledge workers and the economy. The basic premise was that we are looking at the possibility of millions of jobs being impacted in the next 1 - 2 years.
Most comments were supportive of the position. However, there were definitely dissenting opinions and challenges to the thinking, which is great. It’s critical that we have these conversations in the open, and welcome dissenting points of view.
I wrote that post within the limitations of LinkedIn‘s character limit on my ride home from spring break. I knew that there was a lot more to say, so we made it the focus of the Marketing AI Show podcast this week.
See below for a summary of the key points, and check out the YouTube video for the full knowledge work segment.
There are ~132M full-time workers in the U.S. ~100M (76%) are knowledge workers (i.e. people who “think for a living”). This includes writers, marketers, salespeople, programmers, architects, engineers, lawyers, etc.
It is very important that we create awareness for the potential that knowledge work is about to be disrupted.
I understand that people can have a visceral reaction to what appears to be hype. And there are definitely some media and influencers who take extreme positions, and use extreme messaging.
But, if you can sift through the hyperbole, there are some very real issues facing us that business leaders, education systems, government leaders, and professionals are not ready for.
When you consider the current state of AI, and what appears to be coming next, I would much rather industries and organizations work under the assumption that significant job loss is imminent, and be wrong, than do nothing and be right.
1) Capitalism. Companies are driven by efficiency and profits. AI enables both of those, potentially on a level we’ve never seen in terms of speed of transition.
2) The economy is still struggling. Companies are highly motivated in the current economy to reduce costs.
3) Almost 12 million employees in the U.S. (9% of the total full-time workforce), work for private-equity-backed businesses, according to the American Investment Council. A 2019 study by the National Bureau of Economic Research, researchers analyzed almost 10,000 debt-fueled buyouts between 1980 and 2013 and found that employment fell by 13 percent when a private-equity firm took over a public company. Employment declined by even more—16 percent—when private equity acquired a unit or division of a company. AI may be the greatest cost-reduction tool these firms have ever had access to.
4) There is no clear path to government oversight or regulation, and no unified efforts by businesses or educational systems to address the impact on workforces and students.
5) Company leaders are generally unaware and unprepared for how prevalent this technology will be within knowledge work by year’s end. Just look at plans for Microsoft 365 Copilot and Google Workspace. Even if you don’t seek it out, AI tech will be infused into workflows across every business function.
6) We know from our own State of Marketing AI research that the vast majority of companies do not have AI-specific internal education and training programs. In fact, 67% say lack of education and training is an obstacle to adoption, and 81% do not have AI-focused education and training.
7) The tech is insanely good, and largely untapped in its potential. Just look at GPT-4 (language), Midjourney and Stability AI (image), and Runway (video), Replit Ghostwriter, and Github Copilot (coding). Then consider the ecosystem of thousands of apps and plugins being built right now that extend the capabilities of the foundation models.
8) Generative Pre-trained Transformers (GPTs) appear to be General-Purpose Technologies (GPTs), at least according to a March 2023 research paper from OpenAI, OpenResearch and University of Pennsylvania. In GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models, the authors state: "We conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications."
9) It is reasonable to assume the time to complete tasks such as writing, design, coding, planning, etc can be reduced on average 20-30% at minimum with current versions of generative AI tech.
10) This is not the peak. It’s actually very early in understanding and adoption of current forms of generative AI. This tech is getting faster and smarter, so these percentages will only rise.
1) Government regulation. (low probability in the near term)
2) Massive legal precedent that challenges how the foundational models were trained, thus slowing down progress. (likely to happen, but low probability of halting progress)
3) AI creating more new jobs in the near term than are currently being envisioned. (low probability)
4) Companies rapidly adapting to reskill / upskill their talent and redistributing their time. (low probability, but best chance)
5) Adoption rates in companies continue to lag behind the technology's capabilities. Understanding and adoption by industries and professions is the greatest variance to consider, and will follow the Law of Uneven AI Distribution. (medium probability)
6) The flaws and limitations of generative AI are greater than are being discussed in the media, and will prevent mass disruption in the near term. (medium probability, but, not evenly distributed by use case)
1) Focus on education and training for yourself, and your teams. Everything starts with understanding. We offer a free Intro to AI for Marketers class every few weeks that could be a simple starting point. More than 10,000 marketers and business leaders have registered for the class since it launched in November 2021. You can also follow along on AI trends and news with our weekly podcast and newsletter.
2) Create an internal AI Council that is charged with developing policies and practices, and considering the near-term and long-term impact of AI on the company across all functions.
3) Develop Responsible AI Principles and Generative AI Policies in your company. Here's a great example of a Generative AI Policy from WIRED Magazine.
4) Conduct an impact / exposure assessment for your teams. Forget the macro-economic reports for a minute, and focus on individuals and teams within your company. How intelligently automated will each person's job become in the next 1 - 2 years? How does that change your team structure, staffing needs, hiring plans, and business strategies?
5) Build an AI Roadmap for your company. Prioritizes AI use cases and projects, and define how to infuse AI across key areas of the business.