Precisely Wrong
The question to settle before you approve the next AI system.
KeynoteWorkshopBoard session
The promise
The board leaves with one written question and a clear idea of the evidence to expect before it is answered. This is a governance session, not a technology briefing.
Who it is for
Boards, executive teams and risk or audit committees about to automate a decision with legal, safety or reputation risk.
What people leave with
- Why a system trained on good-news data gives confident answers that are wrong in the same direction every time.
- What happens when leaders’ unresolved disagreements are built into automated decisions, thousands of times a day.
- The question to ask after an automated decision goes wrong. Most governance rules cannot answer it.
Formats
- Keynote, 45 to 60 minutes
- Board or risk committee session, 90 minutes
- Working session, 2 hours, with your technology and risk leaders
- Panel or moderated conversation, as panellist or as chair
For boards
In South Africa, King V asks the governing body to govern data, information and technology (Principle 10). It addresses AI directly, with clear accountability for the decisions and outcomes of AI systems and clear, understandable reasons for what they decide. This session helps a board settle that before approval, not after a failure. If an automated decision has already gone wrong, FranklySpeaking’s AI Decision Review finds out why.
The thinking behind this talk: read the essay
Speed is not the risk. Speed built on incomplete information is the risk.
Every talk is shaped for your audience, from the examples to the title on your programme.
Have a date in mind?
Tell me about your audience and I will tell you whether this is the right talk.