Read the recently surfacing news about the faulty chatbots that are being invested in by the California community colleges? Let us decode the entire news and find out what it was all about!
The Problem in a Nutshell
What happens when a student tries to ask a bot an academic question, but the answer received is too hazy? Sometimes, nothing. But what if the colleges had invested millions in making and maintaining these bots? That is something really concerning. It recently happened in the three community college districts in California that reported investing a hefty amount in these chatbot contracts. However, the students have reported an unreliable response scheme. It is not always that the bots misbehave; however, one doesn’t really know when the data can be relied upon. This creates a situation of confusion, even after such a big investment. While no solution has been proposed in this matter, this article just wants to bring the readers’ attention to the matter and stay aware.
The Numbers
If someone imagines how much must have been invested, then people got so surprised? Well, the number may come as a total surprise to you. Three community college districts responded, investing a massive amount of $151, 000. On top of that, the Los Angeles Community College District mentions the contracts and amendments approved from 2021 to 2029, with their cost touching $3.8 million. When such are the stakes, things get a little more concerning than they normally are.
What makes these chatbots so important?
Having an automated system that is readily available to all the students for quick problem-solving is a good idea to implement. For that reason, and for their ease of use, having chatbots is a good thing for the colleges. But because many of these chatbots function by pulling data from static, manually curated FAQ lists and institutional websites, they often struggle with accuracy. If the underlying database becomes outdated, or if a student’s inquiry falls outside the specific topics pre-programmed into the system, the bot frequently provides incorrect or unhelpful responses, which then creates chaos. Essentially, the system is only as reliable as the information humans feed into it. If the content isn’t constantly updated or if the query is too complex for the limited database, the bot’s ability to serve students breaks down. This brings another layer of tension.
Possible Solution
While one can think of many things to deal with this, to solve this, colleges should transition from static, manually curated FAQ databases to modern generative AI systems. By integrating Large Language Models that actively scrape and index verified, real-time institutional websites, bots can provide more accurate responses to the students and whoever asks them anything, for that matter.
