Will AI Fuel a Huge "Services into Products" Shift?

As content streaming has disrupted music, is disrupting video and television, so might AI potentially disrupt industry leaders ranging from Alphabet to Meta, just as search and social media were unimagined substitutes for advertising venues of all types. 

It is one thing to predict that artificial intelligence will  automate many--if not most--complex cognitive tasks. It is another matter to predict what entirely-new value propositions could be created, some of which pose the possibility of disruption of firm and industry market positions, unit economics, competitive dynamics, and business strategies. 

In other words, search and social media disrupted revenue models for newspapers, magazines, broadcast TV, broadcast radio, video on demand and linear multichannel video because search and social media became substitute venues for advertising. 

Media Type	Percentage of Total U.S. Advertising
Digital (Total)	65%
- Search	21%
- Social Media	29%
- Other Digital	15%
TV (Total)	19%
- Multichannel Video	10%
- Broadcast TV	9%
Radio	6%
Out-of-Home (OOH)	4%
Newspapers	2%
Magazines	3%

And perhaps one big change will be a shift of “services” turning into “products,” the inverse of the prior trend whereby products became services. 

What, after all, is a bot or agent that provides what a human professional used to supply? Will “customer service” or any sort of “advice” morph from human-provided expert “service” to an AI-provided product or function?

Content producers worry about this a lot, and should. AI threatens to displace acting, writing, composing or editing (professional services) with AI substitutes that are “products” rather than “services.”

It’s the mirror image of the prior process whereby software and other content “products” became “services” (shrink-wrapped software became cloud-based subscriptions; television shifted from over-the-air broadcast to subscription video). In fact, the list of former products that became services is quite extensive:

• Software (Software-as-a-Service)

• Computing power (Cloud computing)

• Music (Streaming services)

• Movies and TV shows (Video streaming platforms)

• Books and magazines (E-book subscriptions and digital content platforms)

• Car ownership (Car-sharing and ride-hailing services)

• Office productivity tools (Cloud-based collaboration suites)

• Data storage (Cloud storage services)

• Photography (Photo storage and editing services)

• Gaming (Cloud gaming and game subscription services)

• Home security systems (Smart home monitoring services)

• Lighting (Lighting-as-a-Service)

• Manufacturing equipment (Equipment-as-a-Service)

• Transportation (Mobility-as-a-Service)

• Communication tools (Unified Communications-as-a-Service)

It Will be Hard to Measure AI Impact on Knowledge Worker "Productivity"

There are over 100 million knowledge workers in the United States, and more than 1.25 billion knowledge workers globally, according to one Anatomy of Work estimate. And “work about work,” including unnecessary meetings, status checks and information searchers occupy as much as 60 percent of knowledge worker time. 

Hence the interest in AI agents that can conduct activities autonomously, presumably eliminating much of that “work about work.” Customer support, regulatory compliance, security and marketing are areas where agents are expected to contribute. 

To the extent we can measure knowledge worker productivity--and there is an argument to be made that we actually cannot measure it effectively--efforts to boost knowledge worker productivity since 2019 have been quite minimal. 

Title	Date	Publisher	Key Conclusions
How Do You Measure Knowledge Worker Productivity?	N/A	Serraview	- Outputs are intangible and difficult to define - Results often based on team output rather than individual - Companies not necessarily tracking hours for salaried employees - Time spent working increasingly blurred in mobile workforce
The Best Way to Measure Knowledge Worker Productivity	2022	Maura Thomas	- Quantitative metrics don't help in short term for knowledge workers - Qualitative metrics matter most in short term - Best measure is questioning how employees feel about their work
How to Measure Employee Productivity in the Workplace	2024	Robin	- Knowledge work is intangible and difficult to categorize - Existing productivity measures rooted in 'machine age' organizations - Impossible to come up with single universal measure for knowledge worker productivity
Knowledge Worker Productivity	N/A	mediaX at Stanford University	- Serious productivity gap exists between available knowledge and how it is used - Many enterprises fail to fully engage energy and intellect of employees
Research: Knowledge Workers Are More Productive from Home	2020	Harvard Business Review	- Knowledge workers' inputs and outputs can't be tracked like other workers - They apply subjective judgment to tasks and decide what to do when - Can withhold effort often without anyone noticing
"Boosting the Productivity of Knowledge Workers"	2006	McKinsey	Measuring productivity is complex due to the amorphous nature of tasks. Performance is constrained by physical, social, and contextual barriers. Interactions account for a significant portion of work, making targeted improvements essential.
"Measuring Knowledge Worker Productivity: A Taxonomy"	2004	Emerald Group Publishing	Identifies a lack of universally accepted metrics for productivity. Proposes a taxonomy categorizing productivity dimensions and highlights critical areas for future research.
"Rethinking Knowledge Work: A Strategic Approach"	2021	McKinsey	Structured approaches (e.g., workflow systems) improve some metrics but can undermine autonomy and collaboration. Freeform and creative aspects of work are harder to quantify.
Knowledge worker productivity: is it really impossible to measure it?	2021	ResearchGate	Argues that traditional productivity metrics are inadequate for knowledge workers and proposes a new methodology based on human capital efficiency.
Broken Speedometers: Quantifying Knowledge Worker Effectiveness?	July 1905	VeraSage Institute	Highlights the challenges of measuring knowledge worker productivity due to the intangible nature of their work and the limitations of traditional metrics.
What Really Matters for Knowledge Worker Performance	July 1905	Allsteel	Reviews existing research and concludes that a single, universal metric for measuring knowledge worker productivity does not exist.
Measuring Knowledge Worker Productivity	July 1905	Global Workspace Association	Discusses the complexity of measuring knowledge worker productivity and the limitations of traditional methods.

Knowledge workers are those who “think for a living,” making productivity challenging to measure. According to the U.S. Bureau of Labor Statistics, measuring employee productivity means calculating “output per hour” of work. 

But how does one quantify outputs, which often are intangible and difficult to define. Also, when “teams” produce the outcomes, how can individual contributions be assessed?

And there are other complications, such as quantifying “hours worked,” either in-office or remotely. 

None of that stops government agencies from doing their best to measure and quantify knowledge worker productivity. 

For example, total factor productivity in the United States is said to have grown 0.8 percent from 1987 to 2023, but only 0.5 percent from 2019 to 2023, according to the Bureau of Labor Statistics. 

All of which will raise questions when firms and entities start to report “productivity” gains from using AI.  If all we can be sure of is that we can measure or quantify some outcomes we believe to be measures of output.  

Whether that output actually represents knowledge worker productivity is less certain. Most of us would be circumspect about metrics such as “hours worked” or “email volume” or “meeting attendance.”

We’d probably have some greater confidence about tasks completed, revenue generated by a team, assuming it can be identified. 

But lots of common metrics are only quantitative, and cannot measure the quality of work performed or outcomes. People can produce lots of documents, lots of code or “ideas,” but it is hard to measure the quality of those outputs. 

Proxy Measure	Description
Time spent working	Hours logged or time tracked on tasks14
Email volume	Number of emails sent/received5
Meeting attendance	Number of meetings attended or hours in meetings
Task completion	Number of tasks or projects completed4
Revenue generated	Financial output attributed to individual or team2
Information searches	Time spent looking for information15
Internal collaboration	Time spent working with colleagues5
Documents produced	Number of reports, presentations, or other deliverables created
Client interactions	Number of client meetings or calls conducted
Ideas generated	Number of new ideas or innovations proposed

Content Licensing Deals to Train AI Proliferate

As has been the case for earlier generations of conflicts between content owners (media firms, for example) and new types of firms (search, social media), conflicts over the training of large language models is being resolved in similar fashion: licensing deals. 

Microsoft, for example, recently signed a deal with News Corp.’s Harper Collins allowing “select non-fiction back titles”  to be used for training of artificial intelligence models, if individual authors agree. 

The content is said to be for a new model Microsoft is creating, but not intended to “write books.”  

Such deals have become more common as model owners work to defuse content owner objections to AI training using their copyrighted works. 

Content Owner	AI Company	Deal Details	Payments
News Corp	OpenAI	5-year deal for access to current and archived content from publications like The Wall Street Journal, The New York Post, The Times, etc. Includes display of content in response to user queries and sharing of journalistic expertise56	Over $250 million over 5 years56
Various Publishers	OpenAI	Annual licensing deals for training AI models, including companies like The Associated Press, Axel Springer, Prisa Media, Le Monde, and Financial Times56	$1 million to $5 million per year136
The Atlantic	OpenAI	Access to archives for AI model training and collaboration on product development, including an experimental microsite6	Not specified
Vox Media	OpenAI	Access to archives for AI model training and assistance in creating products for consumers and advertising partners6	Not specified
Hearst	OpenAI	Licensing deal for content use in training AI models2	Not specified
Mumsnet, The Center for Investigative Reporting	OpenAI	No deal; instead, these entities have initiated legal complaints against OpenAI2	-
Conde Nast, NBC News, IAC (People and Daily Beast owner)	Apple	Discussions for licensing content archives for AI training, but no public deals announced yet2	At least $50 million over a multiyear period (reported offer)3
Financial Times, Axel Springer, The Atlantic, Fortune	Prorata.ai	Licensing deal with revenue-sharing model; 50% of subscription revenue shared with content creators2	Revenue-sharing basis
Time, Der Spiegel, Fortune, Entrepreneur, The Texas Tribune, Automattic (WordPress.com owner)	Perplexity	Revenue-sharing deal with access to analytics and technology for creating custom answer engines2	Revenue-sharing basis
Reddit	Google	Licensing deal for user-generated content to train AI models4	Not specified

source: Seeking Alpha 

Content Owner	AI/Search/Social Media Firm	Deal Details	Payments
News Corp	OpenAI	Access to current and archived content from publications like The Wall Street Journal, The New York Post, The Times, etc. for training AI models and displaying content in response to user queries. Includes sharing of journalistic expertise155	Over $250 million over 5 years
The Associated Press	OpenAI	Licensing deal for training AI models and developing technology for news gathering45	$1 million to $5 million per year
Axel Springer	OpenAI	Licensing deal for training AI models and developing technology for news gathering45	$1 million to $5 million per year
Prisa Media	OpenAI	Licensing deal for training AI models and developing technology for news gathering5	$1 million to $5 million per year
Le Monde	OpenAI	Licensing deal for training AI models and developing technology for news gathering5	$1 million to $5 million per year
Financial Times	OpenAI	Licensing deal for training AI models and developing technology for news gathering15	$1 million to $5 million per year
Hearst	OpenAI	Licensing deal for training AI models1	$1 million to $5 million per year
Time, Der Spiegel, Fortune, Entrepreneur, The Texas Tribune, Automattic (WordPress.com owner)	Perplexity	Revenue-sharing deal with access to analytics and technology to create custom answer engines. Revenue generated from sponsored related questions will be shared with publishers1	Revenue-sharing basis
Conde Nast, NBC News, IAC (People and Daily Beast owner)	Apple	Discussions for licensing content archives, but no public deals announced yet. Apple is offering more substantial remuneration for broader rights to use the content124	At least $50 million over a multiyear period (reported offer)
Reddit	Google	Licensing deal for user-generated content to train AI models3	Not specified
Mumsnet, The Center for Investigative Reporting	OpenAI	No deal yet. Instead, these entities have initiated legal complaints against OpenAI
The New York Times	OpenAI	No deal yet.  The New York Times is suing OpenAI and Microsoft for copyright infringement