The issue is how you can also easily imagine a lot of negative.
Neither those imagining positives or those imagining negatives can really prove to the other that they're right. This becomes the source of the conflict.
If you predicted the negative outcomes, you'd also oppose it. If you believe the outcomes will be positive, you'd be a proponent.
Now it also happens that a lot of those predicting positive outcomes are currently making money off the development of AI and so have an investment in its continued enablement. Which muddies things even more, because it's an obvious bias.
It's hard to see benefits from AI systems, AGI or otherwise. It doesn't seem to produce anything that improves human happiness or general well being.