Here's an example (actual screenshot, open in new tab to zoom) where a GPT-equivalent AI generated inconsistent and confusing output. Notice that in the AI summary, the content flip-flops between large anode-small cathode and small anode-large cathode being desirable.
The correct concept should be that a relatively larger cathode accelerates corrosion of the anode. As such, a relatively larger anode is desirable in terms of minimizing galvanic corrosion effects.
While an experience will eventually figure out which is corrective. A new learner reading the content as face-value could easily be misled.
In texts related to this topic, the terms anode and cathode appears with equal probability, so it is not surprising the GPT which predicts the next word based on probability will tend to mix this up.
Posts
No comments:
Post a Comment