DL Opinion columns

AI is scary, but more human than you think – by Deborah Levine

originally published in the Chattanooga Times Free Press

Artificial Intelligence is like the weeds growing in my backyard. The AI quickly takes over all the other plants and I often forget that there were actual flowers blooming by my back porch. The impact on our lives is spreading widely. It can be scary when emergency phone lines are now answered by AI. And I was not happy to hear that my son-in-law would lose his position in international banking to AI. So I didn’t know what to think when I was offered a job with an AI company. 

The job offer came just before I got an email inviting me to a course in building my own AI-augmented operating system. I could learn to create: 1) My own Executive Assistant Agent to automate scheduling, notes, and follow-ups, 2) Agents to win business and onboard clients in minutes, 3) Thought Leadership Engine to generate posts, decks, and videos, 4) Analyst to deliver briefs and trend reports on demand. It’s a new norm now to learn this stuff along with an Automation Layer to connect it all into one seamless AI-powered operating system (OS).

But it’s not the norm to be recruited as an experts that AI learns from. We experts are hired to train AI models to be safer, smarter and more reliable. Not surprisingly, it turns out that I had to take an exam before being officially accepted into the corporate community. I decided to go ahead and take the exam because I agreed with the corporate philosophy: “As AI becomes more common, humans must continue to guide its development. Our contributors teach AI systems so that they reflect human values and judgment.”

Yes, I passed the test! 

But onboarding was another matter. There were legal issues to be agreed to, additional platforms to join for access to tasks, and getting paid was complicated. It was a requirement to use an internet device to officially time my work. I would have to meet the stated deadlines, or else. Going through the onboarding process, I have learned more about the internet during the week of setting up than I have in years. Cyberspace has changed evermore than my backyard. 

When it was time to check out my first task, there was an introductory training video that I had to go through first. I may never be the same, especially since I had little understanding of the tasks’ “interleaved generation” topic. It turns out that the topic actually has a well-known abbreviation: RIG (Retrieval Interleaved Generation). Here is the definition of RIG supplied by Shieldbase.ai: “Retrieval Interleaved Generation (RIG) is an innovative technique in natural language processing (NLP) that merges real-time data retrieval with the generation of responses by large language models (LLMs). Unlike traditional methods that retrieve data before generating a response, RIG allows models to continuously access external information while formulating answers, enabling more dynamic and contextually relevant outputs.” 

Most of us probably don’t understand all of that definition, but we do get the fact that life and communication are getting more complex. This means that it’s not just me that needs to learn more about AI and how to humanize it. Our upcoming generations are key to using all this and we should be making sure that education includes AI processes, along with the changes to it that are likely to be weekly, if not not daily. 

I tell myself to ‘Breathe’ and keep going. And I suggest that you do the same. Good luck!

Editor-in-Chief
Latest posts by Editor-in-Chief (see all)