Originally posted by Fogel70 I doubt they designed the camera to support this other than with general firmware updates.
Just for clarity, this is the bit of the OPs post I was referring to:
'It's possible to feed lots of e.g. airplane pictures to train the deep learning algorithm, which could then be released as the new firmware. (Wakashiro san didn't say that that's what they'll do.)'
Which to me implies they could - I'm sort of interested in the would and should
Originally posted by Fogel70 Otherwise it sounds like Minoltas Creative Expansion Card system, where film cameras had a card slot to add features not included in the base camera
Ah yes - I'd forgotten about that - things have moved on though -time to try again?
The only real logic to this is that it prevents even more memory and computing space being taken up by features that individual photographers might not want - but my abilities such as they are are as a photographer, not a programmer or electronic engineer, so I'm really talking on a philosophical level about whether that sort of thing is a good idea,