Great piece on selecting appropriate visualizations from the HBR.
First need to consider the aim of the visualization on two axes. Is the information conceptual (with a focus on ideas and an aim to simplify or teach) or data driven (with a focus on statistics and an aim to inform or enlighten)? And do you intend to declare (with a focus on documenting or designing and an aim to affirm) something or explore (with a focus on prototyping or interacting and an aim to confirm and discover) something?
From this it is then possible to define four types of visualizations:
- Idea illustration (conceptual and data-driven)
- Idea generation (conceptual and exploratory)
- Visual discovery (exploratory and data-driven)
- Everyday dataviz (declarative and data-driven)
- They listen more than they talk
- They don’t always interject their experiences
- They admit what they don’t know
- They are well read
- They look for cues
- They let go of the details
Should we be afraid of AI? is one of the more balanced and considered articles I have read about AI recently.
The article defines two extreme viewpoints that are being perpetuated around AI.
- AI Singularitarians believe that the creation of some form of artificial ultraintelligence is likely in the foreseeable future and that this might involve major risks to humanities future.
- AItheists beleive that true AI is not possible and that there is nothing to discuss or worry about.
The article argues that both viewpoints are extreme and that reality lies somewhere in the middle stating that true AI is not logically impossible but is implausible. What matters is not the possible appearance of some ultraintelligence but the effect that ever-smarter technologies is having on how we live our lives.
The article defines four revolutions in self-understanding:
- We are not the centre of the universe (Copernicus)
- We are not the centre of the biological kingdom (Darwin)
- We are not the centre of rationality (Freud)
- We are not the centre of the infosphere (Turing)
The article concludes:
- We need the smartest technologies to tackle the concrete evils oppressing humanity and our planet.
- We should make AI human friendly in always treating people as the ends and never as the means.
- We should make AI’s stupidity work for human intelligence and that the benefits and costs of AI should be borne by the whole of society.
- We should make AI’s predictive power work for freedom and autonomy.
- Prove it – Demonstrate that the concept has a market. This step is really looking at the product – market fit. Starts with one or two key features and test each with research to see how they resonate with the market.
- Plan it – Plan the development process and product launch.
- Price it – Before designing the product it is important to define the market price to ensure competitive products and margins. The price will then determine material selection and key design criteria and also influences what product features are kept and what features are removed.
- Prototype it – Design and prototype the product.
- Protect it – Provisional IP protection. Ideally you should wait as long as possible before you start revealing details.
- Predict it – Sales forecasting of demand and product longevity in the market.
- Product it – Make it real. For many retail products it is important to ‘babysit’ the producer to ensure the product is produced as expected.
Here is an interesting perspective on oil markets even if I disagree with the conclusion – I personally think US shale oil producers would/can respond very quickly to price increases to balance demand and prevent excessive increases in price.