‘Humanistic’ AI-powered Business

Prelude – The ‘S’ Curve of Business

Every company—regardless of business type or business model (i.e. product, SaaS, service, media, etc.)—will go through a highly predictable cycle of growth and maturity called the “S Curve of Business” summarised well by the infographic below (ref: RocketSource).

The “S” Curve of Business. Courtesy of RocketSource.

As we’ve seen time and again, impressive growth in the short term may be achievable for many companies, but sustaining that momentum without facing moments where growth stalls simply doesn’t happen. The market fluctuates. Competitors make adjustments to stay relevant and spearhead new campaigns targeting and drawing on your customer base. Your ground-breaking product is no longer as impressive and unique as it used to be.

Put another way, from the point of view of transformation and business change,

  • A project typically starts with high energy and limited resources. Think Tuckman’s “forming–storming–norming–performing” model. 
  • Every investment for growth reduces profitability until anticipated benefits are realised.
  • Lot’s of projects and companies fail because they can’t afford to fund the dip. This often includes sustaining backfill resources to compensate for people seconded from the business to help support change. Typically, this includes hiring contractors. However, cross-training employees or having an employee referral platform in place are also options — regardless of which model is chosen, there is a cost.
  • Eventually the organisation reaches peak performance and then requires the next level of investment to go to the level beyond.
  • Add another project … and add another ‘S’ Curve. And the cycle continues. For some companies, this may feel like “boom and bust” cycle with alternate periods of high and low levels of profit and loss.
  • According to InspiredCEOs, 96.4% of businesses NEVER get past the £1m revenue and 10 staff.

So why spend much of the beginning of this article talking about the above?

In short, embracing AI will require navigating the ‘S’ Curve and sustaining funding, executive sponsorship and Board-level support. Each of these challenges present unique challenges and hurdles on their own; they will distract most companies from focusing on topics relating to fairness, accountability, transparency and ethics and perhaps other social issues relating to availability, affordability, equity, bias, etc.

Taken together, all of these ultimately affect the overall human experience of both

  1. employees, engaging in AI projects to create products, services and solutions, and
  2. customers, expecting a positive “human” experience when using AI-powered products & services.

The biggest mistake a business can make is jumping straight into a major change initiative that is akin to “a square peg, round hole” scenario which may lead quickly to the significant failures. Take note of this Forbes article which catalogues lessons learnt from companies that failed at Digital Transformation.

Introduction

Artificial Intelligence (AI) — along with one of its most popular sub-components, Machine Learning (ML) — is a key technology that is radically transforming our world. Let’s pause to reflect on the differences between AI and ML (picture courtesy of Ronald Van Loon).

Machine Learning is NOT AI

We also experience the convenience offered by AI and ML in our daily lives, for example in the form of intelligent search engines, translation algorithms, navigation systems, on-line e-commerce stores, chatbots that automatically respond to questions and complaints, and algorithms that make recommendations to us or even develop tailor-made products that meet our personal needs. AI (and ML) can also be combined with robotics or unmanned systems, for example in the manufacturing industry. AI owes its huge societal and economic potential to the fact that this key technology can be applied in almost all domains and sectors.

From start-ups and scale-ups to small and medium-sized enterprises (SMEs) or large corporations, AI initiatives are crucial due to their innovative and competitive power.

That said, businesses should consider taking a “humanistic” approach to AI adoption by

  • working together in public-private partnerships to realise societal and economic opportunities of AI.
  • focusing on “Explainable” and “Responsible” AI — to uphold fairness, transparency and equity.
  • focusing on equitable hiring which supports diversity and embraces remote working.
  • establishing AI applications that serve the interests of people and society — whilst avoiding #codedbias and other situations that lead to division.
  • opting for an inclusive approach that puts people first whilst we strive for reliable AI.

But this is not a straightforward task for any business. There are some key areas that every company should consider when embarking on their AI/ML journey, namely,

  1. Investing in an AI-enabled Operating Model
  2. Building (and sustaining) an “AI” Talent Pool
  3. Establishing collaboration structures to support “AI” Product & Service implementation.

In short, there are no shortcuts. Additionally, it is not just about technology. You also need to develop strategies that relate to how your organisation will manage data (to support analytics), people (to encourage collaboration) and processes (to support governance).

Without introducing strategies that address the areas listed above, your company may end up trying to fit “a square peg in a round hole” — which will ultimately affect your bottom line and allow your competition to “leap frog” ahead of you.

AI-enabled Operating Model

For any company, the “Operating Model” is HOW a business runs itself. It is how it uses its people, processes, data and technology to execute its “Mission and Vision” and bring value to its customers which may extend to its employees and workforce.

In contrast, the “Business Model” is a strategy that a company sets to drive how it will create and grow value, and WHAT it shall do to achieve it. It is a model used in strategy and planning. It fails when it doesn’t achieve its projected targets such as a particular revenue stream or profitability.

In today’s world of big data, large corporates have access to a wealth of information about their customers, as well as about how they do business and deliver value to their customers. By using AI and ML to build technology that can model and predict these customer and user journeys, or logistical challenges, companies can really leverage their data to find new sources of value.

Building new technology that streamlines a company’s logistics and powers its Operating Model is valuable. When a company has AI assets that can be licenced and sold to other businesses whilst also serving its own challenges, the value (i.e. revenue and profit) of the company has the potential to increase.

Types of Operating Models

When embarking on your AI journey, there are essentially THREE types of Operating Models that should be considered.

Each type needs to be underpinned with a data capability that most likely will need to be established in your organisation; this will be very different from building a digital capability. Many organisations are mistakenly shoehorning AI initiatives into digital transformation — which, in my view, may be a bit short sighted given they are very distinct topics.

1) Centralised model

Centralised AI Operating Model

This model will underpin your analytics strategy. The AI and analytics talent is unified and located in one centralised department that behaves like an internal consulting firm for the rest of the organisation on topics related to AI and ML.

2) Decentralised model

De-centralised AI Operating Model

This model pushes embeds the analytics team in individual lines of business (LoBs); working very closely with SMEs (Subject Matter Experts) or potentially act as SMEs themselves. There might be an IT (or R&D) department centrally working with them, or the IT team might also be inside the business unit.

3) Hub-and-Spoke model

Hub-and-Spoke AI Operating Model

This model aims to provide the best of both worlds, looking for a way to get the benefits of having a centralised team, while keeping analytics talent embedded within the business. It has one central team working in coordination with the AI and analytics talent that is purposely “scattered” throughout and across the organisation.

AI initiatives need more than a “Competency Centre” (CC) or “Centre of Excellence” (CoE) which both require new skills, organisational design structures, and executive support. Being able to adopt AI in a business starts with choosing the operating model coupled with clear business objectives, budgets and real-world use cases — and, dare I say it, upholding “humanistic” values and priorities as part of the operating model so that it is stitched into the fabric of new ways of working.

Whether your operating model is centred within a line of business (LoB) or centred at the heart of the organisation, the people charged with its success must ensure AI and analytics-powered products and services are created with “humanistic” values embedded throughout the design and development process rather than bolted on as an after-thought.

The “Humanistic” AI Team

Globally, companies are adopting artificial intelligence to solve their business problems to become AI-driven enterprises, however, the route is full of challenges/obstacles like

  • No “pipeline” of business-driven use cases to feed pilots or prototypes
  • Lack of data science resources
  • Losing valuable time with manual processes
  • Being not able to trust in artificial intelligence processes

Whether an AI team is an outgrowth of an existing analytics team or an entirely new group, there are many different activities that it can and should pursue. Some of these — like developing AI models and systems, working closely with vendors, and building a technical infrastructure — can be done in collaboration with an IT organisation; others will involve working closely with business leaders.

Building “Humanistic” AI Talent

Demand for skilled AI professionals is a major priority for countries like — the US, China, India, Israel, Germany, Switzerland, Canada, France, Spain and Singapore — in that order.

The US at 100% penetration of artificial intelligence skills, is the benchmark for other countries like China with about 92% of AI talent found here. India, Israel and Germany come close third, fourth and fifth with 84%, 54% and 45% of AI talent found in these countries.

The US stands a clear winner in terms of skilled AI professionals across various categories including research, followed closely by China at number two and the UK at number third. Australia and Canada are the other two countries with highest penetration of AI talent at number four and five respectively.

One of the most critical factors for a business looking to acquire and develop an AI Competency Centre (CC) is recruiting, attracting, or building talent. It is no secret that leading-edge AI engineers and data scientists are difficult to hire — even in Silicon Valley. Most organisations will require a few people with the ability to develop and implement AI algorithms—say, a Ph.D. in AI or computer science. But many of the business-focused tasks of a CC can be carried out by graduates and MBA-level analysts who are familiar with AI capabilities and who can use automated machine learning tools. Alternatively, to get started, businesses may consider hiring consultants or vendors to engage in the early stages of AI projects. In this case, these resources should be combined with internal employees to enable a level of “shadowing” and “on the job training”.

Companies should start now in building and nurturing AI talent. This should not be limited to quantitatively-oriented employees but rather all employees who express interest in “career pivot” opportunities; checkout this link to the excellent work that Sudha Jamthe from Stanford Continuing Studies and BusinessSchoolofAI is leading.

Additionally, more and more companies are partnering with universities to provision employees with access to modular “block programmes” in AI and Data Science disciplines leading to certification as part of Continuing Professional Development (CPD) initiatives. Read more about this in one of my recent postings available here.

Organisational Re-design .. Enabling Better Collaboration

Since AI talent is scarce, it is difficult to develop critical mass if it is scattered around the organisation. Regardless of what Operating Model our business settles for, the organisational structures should incentivise (and reward) cross functional collaboration that is baked into roles and responsibilities.

To avoid excessive bureaucracy, a centralised group should embed or assign its staff — at least some of them — to business units or functions where AI is expected to be common. That way the centre staff can become familiar with the unit’s business issues and problems, and develop relationships with key executives. Rotational programmes across business units can improve knowledge growth and transfer. As AI starts to become pervasive, these embedded staff may move their primary organisational reporting line to business units or functions.

As with many technologies today, AI projects are best conducted in an “agile” fashion, with many short-term deliverables and frequent meetings with stakeholders. If there needs to be substantial system development or integration, more traditional project management approaches may come into play.

At a minimum, the following key roles should be included in every AI Team :-

AI Engineer Roles

  • Focus on engineering and problem-solving skills (not “research” driven);
  • Create technology and data architectures that scale;
  • Create AI applications, products and services;
  • Practice “Humanistic” values (i.e. ethics, bias, equity, etc.) when designing, developing and testing algorithms, applications, products and services.

AI Data Governance Roles

  • Adapt existing Governance mechanisms, policies and processes to be “AI” friendly;
  • Engage in the selection of data sources used to address an AI question or problem;
  • Ensure data sources are representative and uphold diversity and inclusivity principles;
  • Monitor how data is used in algorithms in “humanistic” ways;
  • Assess data quality (and subsequently, oversees data cleansing initiatives);
  • Review and approves data usage by all technical / development / operational teams.

AI Translator Roles (i.e. “business partners”)

  • Has an excellent understanding of the organisation, its strategy and customers.
  • Member of a business function e.g. Finance, IT, HR, Legal, External Relations, who act as an internal liaison or translator, linking functions and business units (at all levels including the “C-suite”);
  • Help a business area to deliver its product and/or service strategy by bringing together the right expertise;
  • Scan AI solutions in the market and used by competitors;
  • Questions and challenges others to ensure “humanistic” values are upheld when designing, developing and deploying AI applications, products and services.

Business Leader Roles

  • Work closely with AI Ethicists
  • Ensure the business’ values, principles, codes and culture are aligned with an ethically and socially responsible business operation;
  • Help protect the brand and reputation of the business by working to prevent potentially unethical designs and bias in products, and consumer backlash resulting from the design and application;
  • Manage any negative legal consequence and liability as a result of biased algorithms — in a “humanistic” way by taking proactive steps to learn lessons and adjust any internal policies and governance frameworks;
  • Prevent any negative financial impact on the business due to one of the above and a resulting decline in market share. 

AI Ethicist Roles

  • Work closely with all of the above roles
  • Advocate against algorithmic bias and unethical behaviours that impact AI products & services;
  • Protect against unintended consequences of AI (via policies & processes);
  • Advise on the impact on consumers of AI applications, products and services;
  • Establish AI frameworks & policies that uphold company standards and codes of ethics.

Ultimately, all members of the AI Team should think and act as “stewards” of AI algorithms, applications, products and services.

Concluding Thoughts

Most artificial intelligence focuses on mimicking or exceeding the cognitive intelligence of humans often lacking any notion of emotional intelligence or “humanistic” values.

Typically, ethical and legal dilemmas arise because of their dominance, combined with ownership by a few big companies. They have assumed essential social and community functions and now set norms in society. From Trump’s screeds on Twitter, to Greta Thunberg’s warnings on Facebook, to YouTube performances by the poet Amanda Gorman, social media is the primary platform for information, free speech and artistic expression.

If your business wants to embrace AI and embark on its own journey towards intelligent systems and automation, you need to create a vision for AI in your company. This will inevitably require new business models and strategies. Otherwise it may be sub-optimal and impact your business, your culture, profitability, and future growth.

Starting your AI journey requires significant commitment – embarking on a change programme, disrupting “ways of working”, migrating to new operating models all while developing skills and initiatives that build long term value.

To build a great house you start with an architect. You imagine what’s possible and then you make plans. You don’t dig a trench and pour concrete into a hole and then think about where the walls should go, what kind of windows you like, what kind of roof you like.

To do it properly, you imagine it. Design it. Calculate the materials. Map out the plans. Hire the best people for the project etc etc. There is an order to success and much of it can be expressed in numbers that help you work it out.

It’s the same with your company when initiating your own personal AI journey.

Start with the end in mind (ref: Stephen Covey) and plan backwards to build forwards.

Key Take-Aways

  • Being specific on what kind of AI capability you want to build or harness.
  • Be clear about what you want to achieve and when.
  • Take your time. Hurried implementation will result in pitfalls that may ultimately cost you more time, money and effort.
  • Companies don’t have to do this on their own – learn to delegate (internally AND externally), and get people in to do stuff you are not good at.
  • Focus on moving from the “here and now” and beyond that.

Above all .. enjoy your personal journey building a “humanistic” AI-powered business.

Reach out to me if you would like some advice, guidance and/or support to get started.

About the Author #aboutme

Over the past 25 years, Salim has built a career in consulting, working both client and supplier side as an interim CIO/CTO and a Business Change / Transformation Consultant.

Salim has engaged in, and led, digital and technology transformations and programmes involving rescue & recovery (“turnaround”), process optimisation & improvement and organisational change — globally across the UK, Central Europe, Nordics, Turkey, UAE, US, Asia and Australia.

Salim is an Oxford University alumni and an author in the field of Artificial Intelligence. Key interests include the role of AI for the betterment of people and society.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s