Artificial Intelligence: Early Economics

Array ( [0] => WP_Term Object ( [term_id] => 1405 [name] => Business Transition Posts [slug] => business-transition-posts [term_group] => 0 [term_taxonomy_id] => 1406 [taxonomy] => category [description] => [parent] => 0 [count] => 55 [filter] => raw [object_id] => 581098 [cat_ID] => 1405 [category_count] => 55 [category_description] => [cat_name] => Business Transition Posts [category_nicename] => business-transition-posts [category_parent] => 0 ) ) 1

In Business Transition Posts

Synopsis

  1. Technological advances generally and artificial intelligence (AI) specifically are developing rapidly.
  2. Those things are highly likely to directly and indirectly alter the long-term viability of many businesses – perhaps in particular smaller ones.

This commentary discusses an article titled The new spring of artificial intelligence: A few early economics. It reports on a survey of 3,000 C-suite executives drawn from 200 companies in each of 15 industries located across ten major countries. Written by two McKinsey & Co. partners, it discusses the development and implementation of AI to date, and where AI may take us all.

Business owners are not exempt from uncertainty and fear – for their businesses, for their families, for themselves. Many will have to make hard business transition decisions in coming months and years. Being equipped to make intelligent transition decisions becomes ever more important and urgent.

Overview statements from the referenced article

“To put it simply, the Industrial Revolution was about machines enhancing human muscle power. The AI revolution is about machines enhancing human brain power”.

The article goes on to report that only five years ago machines correctly recognized images 70% of the time. This in contrast to humans who in 2012 (and presumably pre-2012 and currently) correctly recognized images 95% of the time. Currently, as a result of improved algorithms, data, and computer speed machines correctly recognize images 96% of the time, and counting.

If nothing else, that computer image recognition speed has increased exponentially by 37% (96% is a 37% gain from 70%) in only five years presumably speaks volumes to the ongoing speed not just of AI but of technological advances generally.

Artificial intelligence: who is investing

The article reports:

  1. both venture capital and private equity have tripled their AI investments over the past three years, and “are now investing billions in AI”. This where venture capital investment in AI currently is said to be growing faster than VC investment in biotech.
  2. corporate investment in AI currently is three times that of venture capital and private equity.

Artificial intelligence: awareness summarized by six distinct AI technologies

The article identifies six distinct AI technologies or subsets. These are:

  1. computer vision.
  2. language.
  3. machine learning.
  4. robotics.
  5. robotic process automation.
  6. virtual agent.

Many people – including me – may not know what either “robotic process automation” or “virtual agent” is. This is what I have been able to discern.

  1. Robotic process automation enables the configuration of robot software to interact with other digital systems to capture, interpret, process, manipulate, share data and trigger responses.
  2. Virtual Agent is a term used to describe a computer-generated animated online customer service representative.

The article includes a chart summarizing survey results that break out the following five levels of awareness of each of those six AI technologies as to whether they are:

  1. adopted at scale.
  2. adopted but not in a core part of the business or at scale.
  3. in an experimental phase.
  4. aware of, but not being used.
  5. not aware of.

Of note, 97% of survey respondents were not aware of robotic process automation and 42% of respondents were not aware of virtual agents.

I see this as a chart that begs to be updated in each of the next five years, thereby importantly determining the speed of change that occurs.

Artificial intelligence: adopters identified by technology

Of the 3,000 businesses surveyed 20% were found to be “serious (AI) adopters, 40% were found to be “experimental or partial (AI) adopters, and 40% were found not to be AI adopters.

Interestingly, the non-adopter group is said to be interested in AI, but faces a mix of commercial and technical obstacles – not the least of which is that 28% of firms (which I assume to mean 28% of all businesses surveyed or 70% of all non-adopters) “don’t feel they have the technical capabilities to implement” AI.

The article includes a second chart that for 14 industries and one “other industry category” shows survey results for the “number of AI-related technologies adopted at scale or in a core part of the business” as a percentage of businesses within each industry category.

Business owners in the retail, consumer packaged goods or transport and logistics industries, or advisors to such business owners might want to link The new spring of artificial intelligence: A few early economics. The level of early AI adopter involvement in each of those industries is interesting.

This is also a chart that I think begs to be revisited over each of the next five years to determine by how much, and at what speed, the percentages shown change.

Artificial intelligence: adopter characteristics

When discussing adopter characteristics – and I think this may be particularly important to owners of smaller businesses and their advisors – the authors concluded among other things that:

  1. “adoption is more visible among larger companies” and companies “that have invested in big data and cloud-based architecture”.
  2. investors in AI are companies that rely as much on growing revenues as they do on cutting costs.

Artificial intelligence: productivity and employment

The authors identify two important issues regularly raised by economists (and others):

  1. The likely impact of AI on productivity growth, and whether any enhanced productivity growth from AI can (perhaps the better word is “will”) restore “total factor productivity growth” in developed economies.
  2. The impact on employment AI will have on labour – read “employment”.

With respect to:

  1. productivity, the authors say – paraphrased – that in the long-term productivity growth is a function of profit deployment and expansion of output. They conclude that the total relative productivity change between businesses adopting AI and those that don’t “could be large, and could be even higher … if economies of scale can emerge out of AI”.
  2. employment, the authors express optimism for “future employment”. This where they clearly say that adoption of AI is driven (and presumably will continue to be) driven by economic forces that in turn “will further accelerate the substitution of (old) human tasks”.

Conclusion: things to think hard about

I suggest business owners and their advisors consider:

  1. whether going forward AI and other technological advances – as a generalization – favour larger businesses over smaller businesses. This where larger businesses typically have access to greater financial resources, which in turn give them superior access to capital spending, personnel, expertise, and experimentation.
  2. as a second generalization, if productivity is a function of profit deployment and output expansion, in the end both as a practical matter have to be dependent on consumer spending. This where consumer spending in developed countries typically is reported to account for about two-thirds of gross domestic product.
  3. that it is easy to argue that employment levels, average population income levels, and consumer spending, in the end, have to be inter-dependent to a substantive degree. Accordingly, in the face of ongoing technological advances the question of whether potential net unemployment is – in both social and economic contexts – more important that is the question of increased productivity.
  4. whether it is too early in the game for the authors of this article, or anyone else, to do other than recognize that net unemployment related to technological advances may be a potential consequence of technological advances.
  5. that studies and commentaries on AI invariably discuss productivity gains. For example, see AI to drive GDP gains of $15.7 trillion with productivity, personalisation improvements (PWC June 17, 2017). That article states, among other things, “businesses that fail to apply AI could quickly find themselves being undercut on turnaround times as well as costs and experience, and may lose a significant amount of their market share as a result”. No mention is made of potential AI related net employment changes, positive or negative.
  6. my continuing view that workplace automation through ongoing technological advances and related employment levels can not be sensibly addressed without a discussion around how mentally agile workers will need to be to qualify for employment in a “new normal technology driven environment”. This is a co-relation that few seem to want to talk about – substituting instead an ongoing litany of the importance of education in coming years.
  7. on an ongoing basis, how these and other ongoing economic, industry and business specific changes ought to – or ought not to – impact business owner business transition decisions and the timing of transition implementation.

Our free bi-weekly Business Transition and Valuation Review Newsletter

We invite you to subscribe to our FREE Business Transition & Valuation Review newsletter. Every second Tuesday morning we publish articles and current news commentaries relevant to business viability, corporate governance, business value enhancement and business transition planning.

Not subscribed to our free Newsletter

Sign Up Now

Think this article might be of interest to others?

If you think this article may be relevant or of help to colleagues and friends please forward it to them.

50 Hurdles: Business Transition Simplified

50 Hurdles Cover

The book 50 Hurdles: Business Transition Simplified, is available here at a price of Cdn$37 (including sales tax where applicable and free shipping). Read book testimonials here.

If you are an owner of a family business you might consider purchasing a copy for every member of your business family. If you are an advisor to business owners you might consider purchasing a copy for one or more of your clients. Discounts are available for multiple book purchases.

Ian R. Campbell FCPA FCBV

Ian R. Campbell is a Canadian business valuation and transition expert. He is the author of several Business Valuation texts and of 50 Hurdles: Business Transition Simplified. The Canadian Institute of Chartered Business Valuators recognizes his contribution to the Canadian Business Valuation Profession through the annual The Ian R. Campbell Research Initiative.

He writes The Business Transition and Valuation Review newsletter for business owners and their advisors. You can reach him by email at icampbell@ircpost.com, or by telephone at 905 274 0610.