For example, chips have flaws. Hardware designers work incredibly hard to verify that their chip designs are functionally correct, because the cost of a correctness bug can be catastrophic. With the proliferation of custom hardware designs, verification of the whole stack is even more important. It's bad from the perspective that now instead of building on top and reusing a chip company's huge investment in a verified product, there is a custom product to which an unknown level of verification expertise has been applied. Doubly so because these chips are designed for approximate workloads, so they cut corners left and right on correctness for efficiency. People should be more skeptical.

Being skeptical is fine, but it would be fair to back such a strong statement up with facts rather than completely ungrounded speculation.

This is not how verification works, generally. You have to prove your chip is correct.

I think his point is that you aren't backing up the statement that the chips are unverified. Nor does your statement that they are built for approximate workloads imply that the chips are unverified or cutting corners.