The biggest business problem in 2023

The single biggest problem facing business now is the accelerating rate of technical change driven by AI (Law of Accelerating Returns.).  Author Ray Kurzweil tells this story to explain exponential growth. The inventor of chess asked the emperor for 1 grain of rice for each square on the chessboard.  One for the first square, 2 for the second square, 4 for the third square, then 8, 16….  Assume we are at square 32 out of 64.  That means we are at 4,292,967,296 grains of rice, or a bit over 57,000 bowls of rice.  Exponential growth is almost impossible to visualize – our minds think linearly. 

Technology is growing exponentially, and the fact that chatGPT just passed 100,000,000 users 2 months (according to Reuters) is a wakeup call for every business in the world.  The rate of change is accelerating.  If you think chatGPT is an outlier, think again.   According to Ted Shelton, Microsoft is going to update Bing shortly, and Google will respond quickly.  The race has started, and the pace will quicken. The rate of change extends to all areas of AI, and GPT is just one of many AI areas that will grow exponentially. Companies in all industries must innovate faster.

In the short term (maybe 18 months), the rate of change will double again.  Imagine 114,000 bowls of rice, and the adoption of the next innovation like chatGPT getting to 100,000,000 users in 1 month.  The challenge is our minds, our business systems and our laws, are based on linear thinking. Exponential change compresses the time to adjust. This creates an existential problem for business. The ability to innovate faster will be the most important distinguishing characteristic of the successful enterprise soon – if it isn’t already.

Adopting AI quickly is challenging

Why can’t companies innovate faster?  Most companies have an abundance of smart talented people, and it’s not a skill problem.  The primary reason companies struggle to innovate is corporate structure and process.  Most companies are hierarchical, and their structure is built on a system of management designed in an era of manual labor and manufacturing.  As Peter Drucker correctly said,  ‘the knowledge worker is key, this requires new management, and we must embrace employee autonomy’.  Hierarchical structures require top-down decision making.  Top-down decision making implicitly assumes that people at the top know more than people on the edges engaged in market.  Team members closest to the market and customers see change as it happens. Most companies are not structured to enable and support knowledge workers to make changes. Change requires a long arduous process.

The processes put in place over time to protect the company make the company slow.  For example, selecting a new technology vendor often requires technical approvals, architectural approvals, lengthy vendor management cycles, legal negotiations, etc.…. The processes in most companies are slow because they are based on old knowledge and backward looking models.  As a result of these heavy processes, it takes many large companies over a year at a cost of over $1M to onboard a new technology vendor.  The intent is good – the company wants to make sure it makes the “right” decision.  However, the result is slow, and ironically the “right” or best choice likely changed while the team was running the internal approval gauntlet.  By the time the process is complete, one or more better technologies have emerged. 

Should we buy AI components like we buy software?

Returning to the concept of the accelerating rate of technical change, here is an example of how a process is slow.  Ten to twenty years ago, it was very important to buy the “best” software and take the time to find the software that was technically superior.  As a result, companies created systems and processes to buy the “best” software. These systems and processes have matured. Most companies apply the same processes to buying AI. However, because change is now so fast, AI buying criteria has to be fast. AI solutions that are modular, and deliver results quickly should score higher than alternatives. Companies need to think differently about how they buy AI SAAS offerings that solve a business problems.

The Future and Composability

AI solutions are now mostly sold as SAAS offerings, and AI offerings (software and AAS) will increasingly get better at delivering a business outcome.  For example, a business outcome could be “write marketing copy”.  Today, copy.ai, jasper.ai, anyword.com, are some good alternatives.  What if one of these products scores “higher” based on some point in time quality metric, but another is very easy to buy, incorporate into a workflow, and decouple.  Which should the organization select?  The organization should focus on the later, the application with the highest degree of composability.  However, in most cases, the organization picks the former – the company that scores the highest based on a quality metric, or RFP response. The lens should shift and the primary decision criteria should be the solution that helps the company innovate faster.

Recently Gartner has published articles on the concept of “becoming composable” and they provide solid guidance.  Gartner builds on concept of modular application design. They imagine business users (or fusion teams) solving problems using composable blocks. Cross-functional teams, or “fusion teams” solving problems is a growing industry best practice. While in-house building blocks is a good concept, the rate of change will likely necessitate using external components in the near term. This concept will likely also apply to human capital in the near term.

Emerging best practices

There’s no simple solution to accelerate change, but some important trends are emerging.  Business-led teams with support from cross-functional teams, and guidance from I.T. are a good organizational model for success. In order to effect change quickly, the team needs support from a senior leader, as there will be “exceptions”, and they will need air cover. Aligning procurement to help simplify and expedite composable solutions will become more and more important.

 Adopting a departmental beta mindset at the atomic or departmental level is another important concept.  This mindset enables continual reinvention and the focus is on constant improvement. Finally, and perhaps most importantly, the business mindset must be human-centric.  Successful adapters of AI will respect and support their employees. AI will be used as an enabling tool and not a replacement for workers.  By creating a culture of continual improvement led by knowledge workers, leaders can improve organizational agility.  The single biggest problem facing business now is the accelerating rate of technical change driven by AI. Solving for this problem is the most important thing leaders can do to innovate faster.