naturally artificial intelligence
I want Artificial Intelligence to solve all my problems, but it will fall short just like penicillin couldn't kill all the bacteria. Like penicillin, it is powerful and proven in some areas and not that effective in others. Without its vocal recognition success, I would be swearing less at my magic puck. Without its success in text recognition, my spam folder would be much smaller. With its potential success in visual recognition, I have high hopes for seeing fewer rubbernecks on the roads.
But there is also another aspect of AI, how artificial it is.
Learning algorithms require training. The trainers come in various forms of statisticians, experts in different fields. There is also a need for training material. While some of the training material is publicly available, like weather data, market indexes, the photos everyone posts on social media, and so forth, the crucial pieces lie on my desk. It is this essential internal data that will enlighten my horizon. It is often scattered around and not pasteurized, so it needs to be harvested and homologized by gatherers, often referred to as data scientists.
Since garbage in garbage out holds in all processes, AI gets its share if the harvesting season is short or the outcome is not in abundance. Constructing more materials to supplement a low harvest season is possible; however, the result could behave like a bubble gum rather than chewing gum.
Expert statisticians can build effective models with well-sanitized data similar to pizzaiolo baking perfect Neapolitan pizza with dough from a secret laboratory located in the Royal Palace of Naples dungeon.
Data gathering is as easy as picking olives from trees against a slope of a hill. Bureaucracy, egocracy, and regulatory compliance would act like real icing on top. The overall process is as clean as removing a popcorn ceiling and patching up all the imperfections the paint was covering.
Once the data is ready, the olive oil extraction process can start. I prefer cold-pressing, but other extraction methods are also available. It is one of the rarest moments statisticians crack jokes.
Unfortunately, the story does not end there. It is not an autonomous process. There is a constant requirement for quality assurance to eliminate impurities, making sure only quality olives make the oil. Bias is useful if you want all the constellations whispering you what you want to hear.
In short, I need intelligent beings to tidy up my desk because I am all over the place. I need smart individuals to design a model and teach my desk to be self productive. And I need a group of bright beings to keep an eye on my desk, so it does not make robocalls or push the red button and start a war, killing me in the process and taking over my place.
All in all, please ping me time to time and blame my desk if I don't pingback.