Let's hope so, because I am not so certain.
104 karma · joined June 30, 2023
Let's hope so, because I am not so certain.
Hate the game, not the player, basically.
it might not be "fair", but forgiveness is healthier for you
so never demo LLMs. got it.
if it were opened to public faking such videos would lose (nearly) all of its power
I got some bad news for you then.
3b1b's main selling point is the extreme level of polish on his visualizations - something that takes a lot of time (money) to develop
the sad part is that it takes extreme luck to make it on yt. i wish educating skills counted for more but unfortunately they don't, really.
sure, some times it's more difficult to suspend disbelief, but on the other hand there is many more details that can be appreciated only by people who know how difficult they are to pull off, and are completely missed by people not in the know.
the fact that GPT correctly represents the initial paragraphs gives off the impression to users that they can rely on it for the rest of the article.
from the perspective of a journalistic publication, the fact that GPT can fool people into thinking they are reading NYT's content when it is in fact LLM hallucination has a non-negligible negative impact on their integrity.
this is separate from the copyright issue, but considering how it can be illegal to misrepresent someone's work and words when i leads to reputational damage, i fail to see how GPT can be completely off the hook for this.
Obviously that depends on country-specific slander/libel laws, which admittedly are quite lax in the US, but in general I could see this leading to problems if unaddressed by Open AI.
As an example, if I were to make a website copying parts of NYT's articles and injecting fabrications into the rest, presenting the entire thing as representation of NYT's work, the court would easily rule in NYT's favor.
Despite the default disclaimer that AI work can be inaccurate and whatnot, the fact that GPT will accurately resemble the first part of NYT's content is problematic, as there is no way for the user to know how or why the rest does not follow that rule.
And that's without even going into voice tuning. We've crossed into that world long time ago and the line is only going to get more blurry.
an LLM could educate you in how to commit crimes, which you would have no idea about otherwise
but crimes in general are a bit of an extreme example in my opinion. a better example of risks of unmoderated LLMs would be something that isnt illegal, like for example, manipulating people.
a sufficiently advanced unmoderated AI could provide detailed, tailor made instructions of how to gaslight, scam, and take advantage of vulnerable people.
and unlike straight up committing crimes, the danger of these would be that there is no legal consequences and so the temptation extends to a way wider group of users (including, and especially, kids).
No question you ask will return an answer