AI Strategies: Hair On Fire Yet?

Length: 3-4 Minutes

Have you read the Harvard Business Review article Why Companies That Wait to Adopt AI May Never Catch Up?

The article makes three main arguments:

1. AI Technology is Mature

Recent advances build on findings from the 1980s. “The mathematical and statistical foundations of current AI are well established”

2. AI Programs take Years

Success requires many different activities such as:

  • Knowledge Engineering
  • Complex System Integration
  • Human Interaction Modeling
  • Evolving Governance Models

3. Followers will Fail

If you wait to see first movers succeed, you will have to sacrifice uniqueness. Winners will probably take all and late adopters may never catch up.

Our opinion

All three premises are seriously flawed. We encourage you to reject hair on fire approaches.

1. The technology is immature

The “best algorithms” race continues.

There have been tumultuous methodological and system engineering changes in AI this decade. No doubt we have made major improvements this period. We still need more major breakthroughs to improve the generality, scalability, reliability, trustability and manageability of modern AI technologies. And it is not clear we know how to get there. All major AI researchers are looking for more major breakthroughs.

See the last section of The Easter Rabbit hiding inside AI for more on current limits.

2. Avoid “moon-shot” projects. Start small and grow

The authors create a false choice when they pose only two alternatives to their readers:

Unless you are employing some AI capabilities that are embedded within existing packaged application systems that your company already uses (e.g., Salesforce Einstein features within your CRM system) the fit with your business processes and IT architecture will require significant planning and time for adaptation.

Wrong! Of course you can go way overboard. The authors claim subsequent implementation cycles take many years. Yes.

So we recommend: Do not focus on “moon-shot” “bet the business style” projects that are “science projects” in disguise.

The authors note:

Memorial Sloan Kettering Cancer Center has been working with IBM to use Watson to treat certain forms of cancer for over six years, and the system still isn’t ready for broad use despite availability of high-quality talent in cancer care and AI.

Executives and clinicians in hospitals around the globe have confided they’re not investing in this system and, in their opinion, there is doubt it will ever take off in clinical care.

There isn’t enough publicly available information on the true status of this project. The final implementation, when and if it appears, may be radically different from the early AI marketing spin for it.

I hope the Clinical Oncology Advisor succeeds but this post is not about hope. It’s about fear. Fear of existential doom that the authors are subtly using. And fear I have that enterprises will be driven to invest in far too many moon-shot type projects.

3. Fast followers can be very successful

The HBR article authors words are imprecise.

By the time a late adopter has done all the necessary preparation, earlier adopters will have taken considerable market share — they’ll be able to operate at substantially lower costs with better performance. In short, the winners may take all and late adopters may never catch up.

Of course late adopters and slow followers (when compared to industry peers) will be at real risk. Most enterprises should:

  • Game the potential existential risks that could emerge
  • Stay aware of what’s going on in their own industry and other similar industries.
  • Evaluate moon-shot proposals from consulting firms and vendors
  • Focus on being fast-followers

Net: Go for small, quick, high impact, low internal disruption wins

Find small steps to take that will deliver solid business results quickly. Then iterate and expand your business impact.

In 5 Easy Criteria To Get Quick Returns on AI Investments we exposed recent research out of MIT on generating a 17 to 20 percent improvement in ecommerce revenue by using automatic machine translation technology to expose an English language site in Spanish.

In that 5 Easy Criteria post, we offered five key assumptions to guide your planning. Seek:

  1. Fast time to implement. Can a viable production instance be up and running in 90 to 120 days? (Avoid massive reengineering projects of all types.)
  2. Low levels of internal disruption and reinvention. Let the results drive disruptive change instead of requiring disruptive change to achieve the results.
  3. Suppliers and service providers with the right business mode. (If a significant part of their business is time and materials consulting, you may be failing points 1 and 2 and taking on unjustified risk.)
  4. Relevant real world experience. (Demand verifiable references – use cases – that are already in volume production. Visit them and dig deeply.)
  5. Revenue enhancement (which beats cost reduction.)

Disclosure: I am the author of this article and it expresses my own opinions. This is not a sponsored post. I have no vested interest in any of the firms or institutions mentioned in this post. Nor does the Analyst Syndicate.

Author: Tom Austin

Clearing the fog and misdirection in emerging technologies world wide. I founded TAG after over 40 years at Gartner, Inc. and DEC. Sought out by the press and executives world wide. In the most recent past, I was Gartner, Inc. lead for Artificial Intelligence research (2012-2018.) Deep background in brain sciences (biological constraints on learning.)

Leave a Reply