Back in the day working for McDonnell Douglas, I might have built up a little AI perspicacity. Yes indeed; in the mid 1980s, in the mysterious blacked-out buildings, observing the development of pilot-flight simulation capabilities inside large physical geodesic domes.
Old Style AI Perspicacity
We trained ourselves, then built and deployed things we called expert systems and neural networks. I’m sure those tools and more go into what is now bantered around as artificial intelligence (AI); the new stuff is magnitudes of order more powerful and sophisticated. And these unprecedented capabilities are not sequestered away in some isolated solid black building. The public-facing pieces of these novel AI systems are wide open, but it seems there must be a big back room out there somewhere.
Our old style use of AI wasn’t all for war machines, but that application was the research catalyst. Even back in the day there was debate over good versus evil applications. We did monitor and forecast river movement and erosion. We also did optimized scheduling of national UPS air shipments. This was me, an eager helper. Who knows what kinds of AI oriented systems have grown up? And you don’t have to have big mainframes to play.
New Style AI Perspicacity
Today, AI systems are open to a wide range of end users for any number of applications (uses). As with any modern app, usage and results are highly sensitive to ease of use and interface. Average Joe’s are trying to get good with the AI tools. And if you are smart to begin with, wow.
I well know a smart young data scientist who writes sophisticated code. In this peer group, the competitive edge comes from using AI to assist with your code development. Generate the essence of the code or algorithm, then run it through AI to fix, optimize and/or test it. Speed-to-market and product quality advantages accrue. To gain such advantages, it largely comes down to skill with the system, being able to feed AI the outstanding parameters and constraints. Almost like code to generate code; maybe that is what it is. Or AI spawning AI? Good thing this data scientist is a good person.
Surprising Book Generated – Idea to On the Shelf – In 30 Minutes
I tried this tool, WordGenie by Designrr, and was impressed. A writer friend and I met in the library today, to let him take a look. After poking around to get the feel, I said we might as well try it. We fumbled for a minute to select a topic. I chose “self publishing”. Upon entering it, we were prompted for parameters or subtopics, which went in quickly. It started chugging (not sure how fast the library connection is). After fifteen or twenty minutes of it coming together (hourglass time), it asked a couple format and output questions. Then it quickly spit out this finished product, the tool’s e-book option. I took my name off of it because I really didn’t write it.
Editorial Footnote: The Socio-Political Challenge of New Technologies
Eager minds will be forever vying to “own” the AI space. Will AI be serving primarily “good” or “bad” interests? Rates of technological change and breadth of the capabilities say it will be controlled by a strong collaborative organization. If deployment proceeds without transparency (business as usual), it does not bode well. Bad actors or AI itself could totally squeeze people out of the control function.
The government of this rich country is failing to protect the people in the streets and can barely keep the government lights on. So, no wonder it’s difficult to entrust the establishment with something so complicated and important as AI. We the people need government to actually do a good job on key new technologies like AI and nuclear fusion, not just letting them run amuck and be consumed by elite interests.