rullelito

Why on earth would the fallback when a prompt is under specified be to do something no human expects?

> There are dozens of stories how registering a business alone can take several months and tons of paperwork.

What does this even mean? You have examples from ALL of Europe? Each country has its own process, and at least in "my" country it is very easy.

Drawback is that it only works if you constantly move your head or device.

LLMs are turning into LLMs+hard-coded fixes for every imaginable problem.

They are making it really hard for me to not just go Steam Deck and Linux.

e.g. how to not be fooled by salesmen.

Without knowing too much about ML training, generated output from the own model must be much easier to understand since it generates data that is more likely to be similar to the training set? Is this correct?

Prompts are 99% of AI products these days.

This is similar in zero ways.

To me this is the best thing about LLMs.

Medium complexity tasks in our training set*

This doesn't work on my phone. What is it?

Wait what, you're surprised by this?

But crypto..

> Why does it have to be somehow the government’s fault

Because they bought it.

> They’re so unable to control costs

I assume you talk about the government here?

> Why not just reboot them? Terminating was faster.

If you don't know why you should reboot servers/services properly instead of terminating them..

Yeah, give me free stuff!

Companies keeping their IP a secret? Weird..

Any tool you download can be used as spyware.

Utilizing Girard-Reynolds constraints on a polymorphic push-down automata fundamentally misconstrues both computational topology and dynamic system semantics..

Do most people care? I just want to eat popcorn and watch a movie.

No it is so much better to panic (and vote green, ofc)!

Experiencing the hottest 12 months in the last 200 years says nothing about the longer trend.

Python could be 10x slower and still be very useful in a lot of applications.

So Apple decides which companies should have your location data? Niiiice

You might not need a dedicated product manager at all companies, but someone needs to embrace and act as it. In some situations I have gladly taken that role, in other not.

How on earth would they patch all the historical blocks?

False, you pay CGT on investment gold in almost all countries.

Its value has gone up in most currencies in most time frames. People don't buy gold to buy groceries, to eat it, or to use it as a tool. They buy it so that 20 or 30 years later it has roughly kept its value, i.e. it has performed much better than holding currency.

Buying land requires maintenance, and probably comes with some ongoing tax implications.

My favorite one:

    d = defaultdict(list)
    
    # Magic
    for i in range(2, 10000):
            for y in i in d and d.pop(i) or [i]:
                d[i+y] += (y,)
    
    # Not sure how to print this in a succinct way
    print(sorted(list(set().union(*d.values()))))