I find it strange how seldom the privacy aspects of LLMs come up. Some people are feeding Google, OpenAI, or Anthropic almost literally every topic they think about during the day. Those companies could just use another LLM to find people likely to commit a crime, and then you have the Minority Report in real life.
Yep. In my areas of expertise I can easily catch it coming up with wrong information, bad calculations etc. So laypeople are probably being led astray quite often.
He's got "tech brain" where he thinks everything either serves tech or is tech. But you are correct in pointing out that he's just complaining about inconsequential cultural trends, probably because he's somewhat self isolating or something.
Eh not really. It's more like it starts explaining to the user that they are very special, and they discovered this unique intellectual thing in the AI slop that no one ever has, then starts convincing them it's mystical or similar. Just feeding the user magical narcissism basically.
I installed this a while back and honestly I almost never touch it. It turns out that for me searching my history doesn't really replace a search engine at all. The built in extractor list is pretty limited and adding them seems like too much of an ordeal for me to bother.
> murder in moderation, meth in moderation, punching grandmas in the face in moderation
I know these are your extreme examples but this is literally how society functions in each one.
Meth in moderation = prescription stimulants for ADHD or narcolepsy.
Murder in moderation = war, capital punishment.
Punching grandmas in the face in moderation = arrests of shoplifters or intoxicated grandmas apparently require physical restraint, lots of possible examples of violence for purported good on youtube.
Each of your hard lines are things that literally happen daily around the world, in moderation.
Isn't the industry's idea that 'prompt engineering' is over and anyone can use this stuff effectively?
There seems to be a literal trap where people are too trusting of the LLM and take its word on code or whatever is being offered instead of reading it themselves.
In the context of the classroom this means teaching discernment more than ever.
>"Tiktok is like meth", would you advocate for moderation of meth?
This is an absurd statement. If someone is trying to talk about the middle cases, redirecting the conversation to the edge in order to dismiss their general comment is not appropriate.
'Edge cases exist' is not a lesson most people here need to hear.
Truisms fail when you get to the edge cases. This is well known and you aren't pointing out some massive flaw in their reasoning when it comes comes to classroom AI use.
> I believed this so strongly that my company built an entire product around this concept. I used to tell folks that "session transcripts were the new oil," that they were more valuable than the code itself.
This is pretty funny because it's about the depth of understanding of every 'AI expert' on Linkedin. People who praise the context window as basically magic have no idea how any of this works.