Everything, except the bleeding edge and situations requiring the context window to be more than 60% filled, are elementary to Opus 5.5 . If you provide the questions completely, you can be very sure that Claude’s solutions will be correct. The world has changed; AI is reliable; we need to adapt.

> The world has changed; AI is reliable; we need to adapt.

Yes, but I think it's important to state that it's only reliable for things that are very well known and extensively written about.

I get a lot of obviously wrong or slightly wrong answers from all LLMs across the board. If you don't double-check you just don't know or you might miss it.

And for the bleeding edge part: My godchild wanted to know the name of a Pokémon from his licensed shirt. One of the first ones, found in every training set. Couldn't get it right, was certain it was the wrong one.

> wanted to know the name of a Pokémon from his licensed shirt.

how did you even ask an AI this question? Took a picture of the shirt, uploaded it, and asked the AI what it was? That's a different class of problem than "answer this graph theory question and include your references: <paste text here>"

?

You don't ask for proofs to print them out, put them in your drawer, and brag around that you own the proofs for these exercises.

You ask Claude for proofs to help your understanding. Understanding is the exact same activity as verifying.

According to current practices, you ask ChatGPT or DeepSeek to verify it.

Turtles all the way down, I guess.

If you don’t understand the solutions, how can you verify textbook solutions? And sometimes they’re wrong, maybe subtle typos, causing hours of grief and feelings of inferiority.

The same way you'd verify a textbook's solution.

(Assuming it's a proof or something - if it's just a number - oh well).