zlacker

I have a pitch for a new LLM that will read your tweet before you send it and warn you if it is offensive. $1000/month, half off if you’re a CEO.

>>janals+(OP)
Why would you charge a drunk CEO less?

>>janals+(OP)
This is not a bad idea. Another one is to have it make you solve some complicated math problem or also not let you tweet when you know you are going out.. sort of like not driving to the bar. You could do that with screen time settings though.

>>throwu+51
Right? That's where you make the revenue in your freemium pricing model.

>>janals+(OP)
Don't need an LLM for that (except maybe to get funding), a plain old classifier would work fine at a fraction of the training & inference costs.

>>throwu+51
Nobody’s buying at $1000/month but “giving them half off” makes them think they’re getting a deal.

What they don’t know is they could just have a file on their computer named “is this offensive (open when drunk).txt” with contents that read “yes”.

>>lcnPyl+O3
Um… there are liters lawyers and PR professionals that make their living with that kind of service. And for a certain personality type… I don’t see the 1k$/month being the relevant factor. I’d bet there would be buyers for that service. Though: Probably not enough to actually sustain the liability insurance that you’d need to deal with the complaints of a faulty product

>>lcnPyl+O3
I have fond memories of Gmail's "no emailing drunk" feature and I swear it saved me a couple times.

I think it no longer works, or, well, I grew up.

>>janals+(OP)
As we navigate the complexities of San Francisco Bay Area politics, it's important to recognize that not all partnerships and collaborations meet our expectations or align with my values and vision. While we strive for unity and mutual success within the San Francisco Bay, there are occasions when differences in approach and strategy with specific individuals or groups, such as with Chan, Peskin, Preston, Walton, Melgar, Ronen, and Safai, can lead to a reevaluation of certain relationships. Our focus remains steadfast on innovation, integrity, and delivering value to our constituents and neighbors. We appreciate the efforts of all who are putting forth an honest effort to improve the Bay Area, including those mentioned, but moving forward, we will be redirecting our energies and resources toward partnerships that better align with our mission and contribute positively to our collective goals.

>>hnacco+G7
It's OK. If the client tries to write a nasty tweet about you, or email a lawyer about you, it will "help" the client, by telling them this is offensive and blocking the action.

Doubly so if you try to uninstall it.

>>nostra+93
Yeah, but why would you make that effort when all you'd need is the thinnest veneer over ChatGPT, and given the proposed pricing would leave plenty of margin?

>>riffra+dc
Mail Goggles - apparently released in 2008[0]. I can't find it, so I don't think it exists anymore.

[0] https://gmail.googleblog.com/2008/10/new-in-labs-stop-sendin...

>>nostra+93
Well you're not going to be able to charge them a lot of money without claiming LLM and you won't get funding, so gotta think about the big picture.

>>lcnPyl+O3
You sound extremely business savvy, looking for business partners?

>>bryanr+Bx
Don't worry, you can say something vague like "it's powered by LLM" which could reasonably mean that the LLM was used during the training phase of your own classifier.

>>janals+(OP)
Won't work for people who write under "altered state of mind" conditions. You'll have to show them a fake page featuring the post, store the actual post in some sort of draft box and ask them if they really want to post it 24 hours later.

>>dns_sn+yD
Or that you asked ChatGPT how to write scikit-learn code.