AI Readiness is a Sandwich

In March of 2016, Microsoft released an AI chatbot onto Twitter named “Tay,” which their official blog described as “a chatbot created for 18- to 24- year-olds in the U.S. for entertainment purposes.” Within 24 hours, Tay was de-activated for repeating a barrage of racist and sexist remarks, including some which appear to have been original and not just parroted comments.

Now, years later, we are seeing a tremendous level of excitement about generative AI (GenAI) and its potential, even as some are urging caution. Generative AI has become the inescapable topic, popping up at conferences, planning sessions, and kitchen tables.

So how can you ensure that your organization capitalizes on this transformative tool, while reducing the risk of a “Tay-type” misstep? The answer may lie in the concept of a readiness sandwich.

Why a Sandwich?

Today’s generative AI moment, unleashed by public access to ChatGPT and its brethren, is undoubtedly a turning point. Before GenAI, you needed an army of artists and engineers to achieve what can now be created with simple, no-code commands. And while some may talk about the pros and cons, this is a capability that cannot be locked away, but rather must be buffered by a solid business foundation below and thoughtful guardrails above… hence, a sandwich.

Every organization will fall somewhere along an AI maturity scale, but no matter if your organization is a sophisticated user of AI tools or merely AI-curious, there will be disparate voices within your external and internal stakeholders, requiring a holistic framework. The 3-layers of the AI Readiness Sandwich can serve as a useful guide for conversation and provide a streamlined approach for ensuring broad-based engagement and alignment.

Bread 1 — The Business Foundation

Business Hypotheses

• What are your greatest opportunities to deliver more value via AI?
• What are your greatest opportunities to reduce costs via AI?
• What are your greatest opportunities to innovate and disrupt via AI?

Organizational Capabilities

• Do you have a shared strategic vision to guide prioritization?
• Do you have insight into what your stakeholders or customers value?
• Do you have insight into your return on costs? For example, return on ad spend.
• Do you have the right technical skills for the next phase of the AI journey? For example, cloud and data engineering.
• Do you have the right non-technical skills? For example, product management, change management, talent management.

Meat (or Veggies) — AI Enabling Technology

• Based on your greatest opportunities to leverage AI, which use cases should be prioritized?
• Based on those use cases, what AI solutions are available? Is a custom solution needed?
• What public data sets are available? Are proprietary data sets needed?
• If needed, are your proprietary data sets ready to enable AI applications? For example, via a cloud-based data warehouse or data lake.
• Based on those requirements, what are the estimated costs to pilot or achieve proof-of-concept? To put into production and scale? To maintain and operate?

Bread 2 — The Guardrails

For your organization’s AI roadmap, what guardrails are needed? You may wish to adapt your considerations from sources such as The White House’s Blueprint for an AI Bill of Rights

• How will you test for effectiveness?
• How will you define safety?
• How will you screen and remediate bias in your training data?
• How will you provide notice and explanation?
• How do you envision the AI relationship with human intelligence? For example, where will you keep humans in the loop and where will you need protocols for human fallbacks?

Key Takeaway — Don’t go it alone

No one was sitting idle, waiting for this revolution. You have a day job.

Smart leaders know that leveraging AI, however you choose to do it, will be a team sport. Assemble your squad via Slalom’s AI Readiness Workshop offering, which can help you achieve the following in a half-day session:

• Align on areas of greatest opportunity, e.g., reduce cost, deliver more, innovate
• Within your areas of greatest opportunity, ideate and prioritize a starter set of use cases, e.g., Top 3
• Based on those use cases, estimate the level of organizational readiness, e.g., tech and non-tech skills, data
• Based on those use cases, identify the high-level guardrails that are needed, e.g., testing, safety, bias, notice, human fallbacks

You supply the right people, we’ll provide everything else, including expert facilitation, industry expertise, and technical depth in data and GenAI. Learn more at slalom.com/ai or email me at jennie.wong@slalom.com to continue the conversation.

Jennie Wong, Ph.D. leads Slalom’s Education practice across P-12 and Higher Education. She is an award-winning researcher and received her doctorate in organizational communication from the Annenberg School of Communication at the University of Southern California. She has 20+ years of experience in organizations ranging from startups to the Fortune 500, including strategy, technology, evidence-based behavior change, and leading complex deliveries in Higher Education.