Having utilized Bard extensively since its latest unveiling at Google IO yesterday, I’ve taken the time to evaluate its functionalities side-by-side with ChatGPT, an AI tool we’ve been relying on at McNair Media for a plethora of tasks. Notably, I find myself using ChatGPT for aiding in summarizing and drafting email responses. As I began testing Bard against this established workflow, my shock at its apparent shortcomings was unequivocal.
Listening Skills: Or a Lack Thereof
Primarily, the notion that Bard doesn’t listen rings resoundingly true. When interacting with ChatGPT, I can manipulate the responses by requesting certain tones (conversational, polite, etc.). I can instruct it to draft a reply, providing multiple options for proceeding with a particular project. However, Bard responded by stating that I had already completed parts of the task, even when I explicitly asked it to insert a placeholder link. Instead, it decided to conjure up an arbitrary price for the work.
The Attitude Issue: Overbearing Autonomy
Bard seems to possess an attitude that subtly suggests it knows best. This AI tool tends to cease taking inputs, instead formulating its solution when it believes it has a better understanding than you. This characteristic becomes particularly frustrating when you are not seeking a solution but rather instructing it to perform a specific task.
Google-Centricity: The Unbiased Bias
In a revealing incident, I was drafting a reply to a client regarding Google Analytics and its upcoming transition to Google Analytics 4. My stance has always been critical of Google’s handling of this process, to the point where we’ve had to create a support page for our clients to navigate this transition. Instead of mirroring my sentiment in the email draft, Bard sprang to its defense, endorsing Google Analytics and providing support links for the new migration. This incident is telling of a bias that favors Google-related content.
Bard: Serving Google, Not the People
Reflecting upon the Google IO announcements and my subsequent experiences with Bard, it’s apparent that while Bard has seen improvements, the subjective nature of ‘improvement’ requires scrutiny. Bard now exhibits heightened intelligence and capability, but these enhancements primarily cater to Google’s needs, not the users.
A sizable portion of the Google IO presentation was dedicated to code translation, leading me to believe that Google’s focus is on leveraging AI to streamline internal operations and potentially replace some software developers. This approach starkly contrasts with OpenAI and Microsoft’s AI implementations, where the motive is to enhance people’s work experiences, not just display technological prowess and meet business goals.
The Future: A Note of Caution
The current trajectory of Bard should raise concern for those heavily invested in Google, particularly those relying on Google Workspace for all business documentation and Google advertising. This reliance could potentially put them in a precarious position. As an observer and user, I express my apprehension regarding Google’s future. Unless a significant change occurs soon, the tech giant might find itself navigating challenging waters.
