not sure how legit that account is, actually. It’s not the one I @'ed - this one was created in Jan 2024 - either it’s his low-key alt or a bot

perhaps his plausible deniability account.

the thing about this is a “writing assistant” doesn’t have to be integrated into the email product. It could be a product in itself. If the “business audience(?) was the most interested in a writing assistant” you’ve got a fucken great standalone product opportunity on your hands. A Proton-certified LLM writing assistant that is magically better and more secure than anything else out there is not something you coyly slip into the email client and nudge everyone to use.

It doesn’t matter what reasons they have for doing this. Their method of deployment says more than enough to me. They know it’s off-script and they know they want their fucken “audience” to do the marketing for them.

Nope it isn’t. You blame AI companies for pollution they didn’t generate. Blame the energy sector.

don’t talk about providing good counterarguments, please

We’re testing SearchGPT, a prototype of new search features designed to combine the strength of our AI models with information from the web to give you fast and timely answers with clear and relevant sources.

this is the most wishy washy distinction from chatgpt as a thing. Blatant throwing-shit-at-us-to-see-what-we-eat

from https://web.archive.org/web/20240726195743/https://openai.com/index/searchgpt-prototype/

fyi they updated their blog post with this catch-all disclaimer in the last couple of hours

“it is simply too big to categorically endorse or not endorse”

“so we’re gonna play it safe and endorse it”

Is it absurd that the maker of a tech product controls it by writing it a list of plain language guidelines? or am I out of touch?

“it is providing Microsoft non-exclusive access to advanced learning content and data to help improve relevance and performance of AI systems”.

I wish it wasn’t normal to call these “systems” instead of “products”

I joined a writing meetup here in Amsterdam which gathers every week in a bar to write, to talk about their writing, to bounce ideas, etc. I kinda got tired of going because there were a worrying number of people using chatgpt to generate ideas. I was the only one trying to write non-fiction, and most of what I was writing would be crit of tech (sometimes genAI) so talking about my writing was always fun. But nonetheless, their use of chatgpt seemed extra weird because we were there, together, to write and support each other, for free.

It’s strange to use solidarity, support, and just general helpfulness from others as an explanation for how AI opens writing up to classes or abilities when that’s probably one of the top things that social media (and pre-social media social media) gave us on the internet.

anyway…

the usefulness of any feature should be measured in how deep you can bury its “opt-in” option in the settings pages without hurting its adoption

Eamonn Maguire, author of the Proton Scribe announcement post, responded to my tweet with this: https://x.com/EamonnMagu14645/status/1814062340863651965

We built this as an opt-in alternative to the non-privacy centric options on the market.

Our goal is always privacy by default, we want to make that possible in the GenAI world too given the number of businesses already using it, and the privacy risks other options pose.