Nice work! How do you verify correctness of the generated exercises and explanations? To me this looks the biggest risk in becoming a user: what if my _teacher_ teaches me subtle nonsense that I cannot easily detect since I'm learning and unfamiliar with the material (even if it's only in the 1-2% of cases)? Human teachers make mistakes too, but an LLM cannot _understand_ that it made one...
So.. how do you solve for this issue of trust?
they dont verify, they just present a LLM app and the user suffers if the information is not correct. however most of the time it is correct but sometimes it is not. one way to verify correctness is to ask a bigger model like OpenAI o3
I’m not sure how the solution to “LLMs lie” is “more LLM”. I’ve personally had o3 tell me stuff like: “I ran this query against version 0.10 of DuckDB and verified that it works” when the query contains functions that don’t exist in DuckDB or “this version gets better PageSpeed Insights results” when I know it can’t check those. It happens surprisingly often and it’s super obvious. However, it’s made me seriously wary of the information that it gives which I can’t verify purely based on logic.
But indeed, if they had the _option_ to escape from reality from time to time, people like Cypher would be happier (I know I am when I do it :D). Indeed the irony of partaking precisely in the thing that you fight against is there; however, what they actually fight against is the enslavement by it, not the casual enjoyment of it: it's bad to be a medieval serf forced to toil the soil, it's quite another thing to play gardener to take your mind off the day..
They have an unsophisticated, binary view of the world (for plot reasons, I guess) while some nuance and tolerance would have had made a more realistic and_human_take on the situation.
Cypher: You know, I know this steak doesn’t exist. I know that when I put it in my mouth, the Matrix is telling my brain that it is juicy and delicious. After nine years, you know what I realize? Ignorance is bliss.
Agent Smith: Then we have a deal?
Cypher: I don’t want to remember nothing. Nothing. You understand? And I want to be rich. You know, someone important. Like an actor.
Agent Smith: Whatever you want, Mr. Reagan.
Cypher: Okay. I get my body back into a power plant, re-insert me into the Matrix, I’ll get you what you want.
Cypher didn't just want to eat fake steak, to him that wasn't enough. Ignorance is bliss. He wanted to eat fake steak and not know that it was fake.
To add to this, in most media that shows virtual reality, people who are addicted to it are also shown. Those people, even knowing it is not real, still enjoy riding roller coasters or whatever. Yet in The Matrix franchise there isn't a single person (that I know of) that is addicted to training programs or some such.
It may depend on how the whole ecosystem evolves: if the competition starts doing similar stuff, then you would have no alternative but to stick with one of them and accept the introduction of ads.
That's the funny thing about oligopoly, you know...
The success of Netflix came as it was more convenient to use it than to download. If all streaming service are not seen as convenient everybody will be back to "old" download strategy.
yeah.. these days you really can't tell; every opinion, its counter-opinion, its meta-counter, its sarcastic spin, its trolling version, its counter-trolling, meta-counter-trolling versions.. everything (in the absence of body language cues etc.) is equally likely, which if I recall correctly is the definition of maximal entropy of information, i.e. random noise
