Data is the lifeblood of much of the world today. Not necessarily big data, and certainly not perfect data, but organizations, individuals, governments, really everyone is making decisions based on data. You might think it's going to rain, so you cut the grass today, or maybe defer adding fertilizer. Your organization sees demand for a product, so it orders more and produces more. Government is always using data to make decisions. We might not think they make great decisions, but data matters.

Recently I was reading a science fiction book (I, Starship) about the future, where a person's brain (Henry) becomes uploaded to manage a starship. This ship will travel light years away for 80 years and they need the human crew asleep in hibernation to survive the journey. The interesting thing, to me, was a part in the book where there is a discussion of why Henry was uploaded and why AIs aren't advanced enough to run the starship. There's this quote: "The first generations performed well, but as time went on, we entered a situation where more and more of the data available to train them on was itself machine-generated. So, instead of mimicking high-quality human output, the outputs got more garbled."

Read the rest of Collecting Data Is Hard

Podden och tillhörande omslagsbild på den här sidan tillhör Steve Jones. Innehållet i podden är skapat av Steve Jones och inte av, eller tillsammans med, Poddtoppen.