Step 1
The 3 key risks around GenAI output that directly impact accuracy, compliance and trust.
What are hallucinations, bias amplification and data leakage?
This GenAI alternative is great for tasks with known inputs and exact, calculable outputs. Its examples include formulas, lookups, data transformations and macros.
What is traditional compute?
These core systems are often used between operations and customers, and could be integral to your company's competitive moat.
What are systems of engagement?
Value risk, feasibility risk, usability risk, and adoption risk are the four categories of these. You prioritize the riskiest one first.
What are assumptions (that must be true for the solution to work)?
This is Step 5 in the Applied GenAI methodology
What is Prototyping and validation?
Your data stays private in this type of AI tool.
What are Enterprise AI tools?
The way GenAI costs are modeled.
What is usage-based
These back-office, employee-facing tools drive business operations.
What are systems of record?
This happens to AI projects that don't conduct early testing.
What is fail?
This is why enterprise AI tools exist separately from consumer tools like free ChatGPT. It's the reason your company pays for a business license.
What is data privacy, governance, and security?
The 6 steps of the GAAI Methodology.
What are Finding the opportunity, To GenAI or not to GenAI, GenAI unit economics, Build, buy or both, Prototyping and validation and Measuring and evaluation?
These 4 types of patterns signal GenAI might be a good fit.
What are time bottlenecks, information synthesis, repetitive cognitive tasks and content generation and quality?
This GenAI solution strategy is employed when you need to customize some new functionality on top of an existing platform.
What is hybrid/both?
This is the purpose of the validation phase.
What is to test core assumptions?
This is the most important thing to get out of the validation process
What is learning?
The curriculum's 4 strengths of GenAI.
What are synthesis of unstructured data, natural language interfaces, generating contextual responses and handling variability?
Precision math, deterministic processes, and structured data lookups are examples of tasks where this is the correct answer to the module's title question.
What is 'Not to GenAI' — GenAI is not the right tool
Reduced differentiation, data ownership, brand reputation and vendor lock-in are all risks of this GenAI procurement strategy.
What is buying?
These are the two conditions when it's appropriate to deprioritize an assumption to validate.
What are high confidence and low impact?
This is why GenAI sometimes gives different answers to the same prompt, and it's a feature to work with, not a bug to fear.
What is non-determinism (or AI variability/prompt sensitivity)
The 2 most common traps to avoid when finding a GenAI opportunity.
What are automating a broken process and the cool demo problem?
This is the reason you can't skip the business case, even a brilliant AI idea dies without one of these.
What is financial justification (or ROI)?
This is the best strategy when you need to automate an internal process that is common across most organizations?
What is buy?
These 3 metrics comprise the 3 levels of measurement that should all be considered, post-launch.
What are component, task and outcome metrics?
This is the first step in a solid measurement plan.
What is a developing a clear hypothesis tied to the opportunity statement?