And I use SOTA models that are hosted on hardware I don't have access to a handful of times a month at this point, and most of the people I talk to regarding LLM usage are either in the same boat, or are asking about hosting things themselves. The diminishing returns for using the best of the best models just aren't worth it for most things anymore.
Because the economic, social, and climate impacts are at best uncertain and at worst devastating to the majority of human beings on the planet. It's always surprising to me that people don't intuitively separate the usefulness of AI from its risks. It feels like everyone has to be a doomer or a booster.
Beautifully put. This exactly summarizes my feelings about this particular tech. I actually find it hugely useful. But I fear we lack the wisdom to use this in a way that won't destroy us.
Pretty much, my car doesn't have navigation built in but supports android auto. I've tried a phone mount before but the experience is so much worse when it's not integrated with the car (volume of announcements and music can't be done independently, can't switch what music/podcast app's playing without grabbing my phone, etc) that I just went back to using it.
Why not? A recording from a robot can be potentially viewed by a real human, if the company is not reputable (happened with robotic vacuums). But by default, no humans will view it.
A human house cleaner by definition is a real human looking at your home every time.
One human, with a single set of human eyes and a limited network of associates with few immediate criminals looking to take advantage (although, yes, that does happen).
Versus a small army of roomba's and equivalents mapping out floor plans and tangentially exposing human occupation times with that large data set almost certainly exposed to criminal and surveillance networks that will almost certainly use some of that choicer data to their advantage.
reply