OliverWyman Forum consultants released 100-page report for Senior and C-level management about how GenAI transforms the modern workforce, society and consumers’ habits. And it is a very exciting reading!
The authors interviewed 15,000 respondents in 16 countries during October-November 2023 and presented the report at Davos 2024 conference. The report provides both high-level picture of GenAI impacts but also go into details into regional, industry and generational differences for the current state of GenAI adoption and potential future.
Apart from the obvious things, like the speed of adoption (17 years it took for Internet to acquire the same amount of users that ChatGPT did in 10 months) and overall optimism of people about AI (96% of respondents of say generative AI can benefit their jobs), there are very interesting insights that were interesting for me to read.
Mass adoption ≠ mass productivity
The dramatic uptake in generative AI has been useful for many but hasn’t yet resulted in significant productivity gains across the board. Why?
Authors suggest that we may face what is called the productivity paradox (also the Solow computer paradox). It is the peculiar observation made in business process analysis that, as more investment is made in information technology, worker productivity may go down instead of up. This observation has been firmly supported with empirical evidence from the 1970s to the early 1990s.
Workforce pyramid disruption
While it is clear by now that entry-level jobs will be affected the most with GenAI adoption, there will be an another effect as well. As generative AI replaces some front-line roles, it will disrupt the pipeline of manager roles, akin to the “collapse of the middle” in the job pyramid. To simplify, the role of a front-line manager/supervisor may become obsolete as this role could be replaced with former “junior” employees.
Disconnect between employees and employers priorities
There is a clear difference between what employees and employers see what is important to learn today. My opinion on this discrepancy is following – AI adoption will happen as part of companies strategy, so employees will learn it how to use AI anyway. It will be just part of roles descriptions. But analytical and creative thinking is way harder to master than prompt engineering.
Only after I pressed “purchase” button, I have realized that this book has been published back in 2018. Feels like ages ago. But it was really interesting to see what authors’ predictions came to reality and what are still just concepts. Many new cool technologies from 2018 did not cross the chasm.
This was the first HBR’s Must Reads book I finished in my life and there are several pros and cons with such format. First of all, different authors have different writing styles and it is a bit confusing for a reader, especially, because quality of writing varies as well. Second, writers are seemed to be aware of very short attention span of readers and try to pack as many ideas as possible and trim the text. It works, but it leaves a reader with a lot of unpacked thoughts.
Out of ten articles the most interesting ones were “Marketing in the Age of Alexa” , “Collaborative Intelligence” and “When Your Boss Wears Metal Pants”. All these articles provide analysis how humans will collaborate with AI in various workplaces and businesses. While we don’t see such collaboration in full yet, there are clear indications of this trend. The breakthrough will happen at that moment, when AI will be considered part of the team and will be participating in tasks assignments on par with humans. The article “When Your Boss Wears Metal Pants” gives overview of some experiments and researches how humans will behave in such situations. The results were quite surprising for me.
There are a lot of predictions that did not become a reality. For example, we don’t see massive usage of commercial drones, while there is a clear use case for them as a weapon. Marketing organizations did not change their objects from humans to AI Assistances and, overall, AI Assistances are very far from the point where we can delegate to them complex tasks and rely on their decisions to make purchases on humans behalf.
Several articles touched how business is and will use AR and AI/ML (like, Stitch Fix) and it is funny to realize that from a consumer perspective you may not know that a product or a service value was generated for you by Artificial Intelligence. As a technologist, I am aware that infusion of AI/ML capabilities into apps and tech products happening on a massive scale, but it is rarely visible for an end user. So, those predictions from 2018 became true 100%.
Overall, this book is 4 out 5. It has several interesting ideas, but they are worth only if you want to learn what was a trail of thoughts in 2018.
During latest Christmas holidays I read State of Phygital Report that covers definition of “phygital”, use-cases and analysis of impact on existing industries and verticals. Below are some my thoughts around it.
I guess, many of us wait for more AR/VR features in our smartphones and consumer electronics. In fact, there are already dozens of features and apps are available today. So, what is the next step? The authors of the report believe that is is a Phygital revolution.
We see Phygital as the philosophy of a new world order, where Phygital essentially enables the close integration of the virtual environment (digital) into real human life (physical).
I have recently came across State of AI report and want to share some of my thoughts around it.
First of all, the report has a lot of data but most interesting for me was Industry part that talks about companies and their products in AI area. I believe it is obvious by now that almost every industry is or will be affected by infusion of AI/ML features into products, workflows and processes.
Some notable examples from the report:
Use of AI-based microscopy to find most effective cancer drug to improve survival
UK National Grid Electricity System Operator has implemented new electricity consumption forecast system that more than doubled precision of forecasting
More than 300 different apps are using OpenAI GPT-3 integrations that currently generate an average of 4.5 billion words per day
And it is not surprising. Almost every industry has to deal with capacity planning, future prediction and forecasting – areas where AI is superior than humans.
Among other apps that are using GPT-3 integrations is Github Copilot which is basically converts comments to the code, can create functions and suggests unit tests. How fast such systems will replace Software Engineers?
And it is not a rhetorical question – in a world-first, South Africa granted a patent to an AI system. The system, called Dabus, invented a method to better interlock food containers. Most countries, however, do not recognize a machine as an inventor.
The patent application was submitted to patent offices in the US, the EU, Australia and South Africa. It was rejected in the US and the EU, and a particular ruling on this patent is still in waiting in Australia. In the US, a judge ruled that only a human can hold a patent, not a machine. This is because according to American law, “a natural person” needs to take an oath that they are the inventor. A contradictory ruling came out in Australia, which stated that an AI can be named as an inventor in a patent application.
Now the question is will we have enough of Critical Raw Materials to meet demand of High Perfromance Computing systems for training and running AI/ML models.
Countries accounting for largest share of EU supply of CRMs