I received an invitation to collaborate on “What do you do if you want to break into the job market as an AI career changer? ” on LinkedIn. I did not understand the question but finally looked at the full list of collaborative articles.
The AI generating these categories, article titles-questions, and answers outlines has been “creative” over the top!
Naturally, I went to explore the Boolean search category. Since Boolean search syntax differs on different sites, I thought that questions in the category should mostly be related to specific platforms, like Google or LinkedIn. It’s not the case.
What LinkedIn AI has generated (italicized, with my comments):
-
How do you design XOR and NAND puzzles that are challenging but not frustrating? What? I’d say it’s a cool question to ask ChatGPT in your spare time, but it is unrelated to Boolean search. It’s about complex logic – not implemented in our practical search systems.
-
What are the advantages and disadvantages of using Boolean strings in job boards and resume databases? When we are talking about disadvantages, I expect to hear about alternatives. Do not use Boolean search to find resumes on databases or LinkedIn, and do what – avoid search operators? But here is what it says:
Using Boolean strings in your recruitment process can be challenging and risky.[…]
A lack of familiarity with the syntax and logic of Boolean strings can result in a steep learning curve and a high margin of error, leading to frustration and reduced search productivity.
I never knew Boolean searches could be “risky” and reduce productivity, but here you go!
3. How do you evaluate the relevance and ranking of web pages based on their titles, URLs, and snippets? That one went to the SEO land, which uses some Boolean searching but differs from “Boolean search.”
4. How do you test and debug your Quine-McCluskey method and prime implicants solutions? I Googled the method name – it is about math functions with only 1 and 0 values, referred to as Boolean. It’s not about search.
5. How do you communicate and collaborate with others using Boolean derivatives? (Do you know the answer? lol)
(There are more interesting questions.)
It looks like LinkedIn’s “collaborative article” AI generates these barely relevant questions with semi-hilarious suggestions based on a large pool of knowledge—too large and without context and proper training. It picked the word “Boolean” (which can mean several things) from the request and went for wide interpretations. In this case, it also gathered irrelevant knowledge from scientific systems or libraries. Some questions I saw seemed OK and relevant, but most were off.
You will observe the same story—unrelated or odd questions and suggestions—with another familiar collaborative topic: Sourcing. For example, it has this question (and I don’t know why): How do you design and implement a fair and transparent competency-based pay structure?
Members who come in to respond, I’d say, act “politely,” try to make sense of the questions, and if a question is about something obscure (like you saw above), even Google and paste answers! LinkedIn promises you will stand out as a collaborator.
I previously wrote AI and LinkedIn Are Like Oil And Water; this post is related. So is this – LinkedIn Wants Me to Pack Groceries and Learn Fluoroscopy.
We have just announced a new hands-on practical
ChatGPT for Recruitment Workshop
on April 23, 2024 Tuesday. Participation is limited to 10 people.