Comment by solatic
14 hours ago
The way the post is written, I wonder if the author is working for a company going through a growth spurt and where, through sheer size, everything is becoming more "corporate".
There's a huge difference between having AI clean up a text you send privately to someone you have worked closely with for years, versus a broad spectrum text sent by a VP to hundreds of people or more. The first case is reprehensible, for the reasons the author lays out. But as for the second case, corporate doublespeak has been a meme since long before the advent of AI and it would remain even in some AI-pocalypse. Just because your boss puts out sanitized language in a mass communication, doesn't inherently mean your boss won't still be present and real with you in a more private setting.
I wonder if the way that LLMs write - the super generic, soulless communication - is also due to them being trained on public communication.
On other side it is what they are tuned to produce. It is what majority wants or accepts or at least expects. If that type of content is everywhere. And various types of "leetspeak" or slang is considered unprofessional and thus undesirable. That is what you have left.
On other side, you have to think uproar what something that could be perceived as say racist by most radical people would cause... You really don't have much left and to aim at blandness. With certain flourishes that makes you look more "Learned".
Claude says it can speak 5 kinds of 1337. It even built a simulated blue box for me complete with keypad, 2600 button, and KP tone button. Then it made a modem simulator with buttons for all the different tones/warbles/screams.
Which is to say, tuning fights with prompts less as the models and interfaces improve.
It absolutely is their alignment on this. They are purposefully designed to be like that.
Yes, the more personal the context, the more the humanity aspect / being relatable matters.