amacalac

They get that many?!

As far as I understand it; Jev et al can be considered Zero Shot Classifiers.

It’s likely the system you built was trained on images of just cats and dogs; but Jev is using a general training set, and there is tuned for general purpose zero shot classification, rather than domain specific “is this a cat or a dog”.

Yes it’s boring, but where did it win?

shrug – same reason Google Meet now looks like a duck...

Can you share the link or has it bean and gone, and skipped your mind?

I’m not certain here, because I think Cursor drives a chunk of cash through token usage to Anthropic.

Show value in as little steps as possible.

If it needs access to my real system data, tell me why, and guide me through the all steps I need to complete to set it up.

Write out every step. Make it obvious. Come back to your instructions the next day and make them simpler, and more obvious.

Have as few steps as possible.

Time to value is key here.

$50 to move past your bottleneck? I’d say that’s worth a fortune in ROI if you had:

a) poor design sense; b) poor design engineering (you have good design sense but can’t figure out how to build it); c) taken more than your hourly salary to learn a + b.

Agree with understanding if you could do it cheaper and faster next time, but the rule of 3 usually applies: “good, fast and cheap, pick 2”

It looks like you’re on the right track; pragmatism about specific tools, and their use cases.

Hopefully your friend sees the value in the tool and a big ROI too!

Now the Ex-X CEO!

Still hoping to grow legs for the Metaverse?

It's not very confidential then, is it?

Doh! Ray! Me!

Yes. As shown in the movie Minority Report, this will surely be flawless in preventing all future crime.

this is fine.