2 Comments
User's avatar
Roger B's avatar

Can you address the potential perception that those who program Bard to populate data on behalf of Google, may carry predilections in favor of a Blue agenda, placing more weight on those accomplishments? I don't claim to know, but much of the disclaimer following the Red accomplishments could also be said of the Blue accomplishments depending on your perspective.

Expand full comment
Edwin Eisendrath's avatar

Good question. Bard's technology is explained here: https://blog.google/technolog y/ai/lamda/). Google makes clear that there are errors in all large language model AI techs, including theirs. But these errors are not likely systematic errors relating to politics. I found a similar result using chat gpt on its own and as it is deployed in Microsoft's Bing.

As with many new technologies, I think it's important to see if the answers pass the common sense test. These responses do. I've been looking at state level politics and government for a long time, and these responses definitely track what's happening.

Expand full comment