I don't disagree you should improve your skills. but I don't think your app is optimized for that.
But, ok you're the gamemaker, then well done building something, and I hope people enjoy it.
For me, an analogue watch is good for giving you an extremely glance-able approximate time. And potentially as exquisite mechanical jewellery.
If you need to know the really exact time to the second and want to be able to tell it quickly, you might be better with a digital watch, especially vs a mechanical watch where a mickey-mouse digital is more accurate than even a COSC mechanical.
But my initial point is that your game is not so much about recognizing the time, even the exact time, but about entering it into the keypad.
Maybe add a multiple-choice mode - show the analogue clocks, pick the digital match.
We've got used to 'reasonable' society and politics in the last few centuries, but check out politics in developing nations or dictatorships, or woke, or Trump or places like pageantry. All fake news and gossip and performance, and AI just makes this potentially much much worse.
It already has a name in academia I think, post-truth, or post-reality or whatever, I think it all started with the French post-modernism thing, then critical theory, etc.
But I'm not sure it's a post- 'advance' at all, but more like a rejection of the enlightenment and a return to the tribal village.
lol, this completely misses the point of analogue watches or clocks.
It should be called 'keypad data entry game' instead.
The point of an analogue clock is that it gives you an instant sense of the rough time at a glance. I can immediately see it's 'about half past' or 'only a few minutes left until the meeting at 11' or whatever.
Semple's humorous long-form performance art, pillorying Kapoor, while developing a menagerie of new pigments, is much more interesting than Kapoor's sculptures.
Anyone could have made at least the same with the given material, and a more talented artist could stun, but sadly they are not allowed to because contract.
It actually does work for a whole range of things.
Is it a panacea? Of course not, but it is potentially a sustainable business.
Except - open models are barely months away from the same performance and there is no moat, so we will see commodity pricing, and the billions being heaped on the fire currently will probably not see a return, unless someone comes up with a genius trading bot perhaps.
The first example only tells me that the energy is dependent on your frame of reference, since the collision seen from the train appears to have more energy than the head-on collision, simply due to the moving viewpoint, whereas they must be the same.
Surely the roots, if we skip over the early preceptron work', are in backpropagation and Hinton, and the work going on at Edinburgh and elsewhere in the 80s.
Indeed I remember buying a set of three conference-papers-as-books around that time, titled Artificial Neural Networks .. proceedings of the whatever the conference was.
No doubt Schmidhuber made important contributions, but I see him pop up claiming to be the 'root' of it all every couple of years.
We should be teaching people to be cynical of AI answers.
Even if the answers are correct, they could still be biased, incomplete, misleading, and all the other media-literacy things people should be looking out for.
This ruling seems to go the opposite direction; 'I am legally obliged to give correct answers, so I am always right, trust the AI'.
IMHO, as an analogy, matter is not 'a bowling-ball on a mattress', but more like a scrunched-up section of table-cloth. Tiny knots or whirlpools of space-time/quantum fields, different particles are different topologies of knot, albeit the nature of space-time is unclear and it may well be a projection.