If you’re having a day where you feel like you’re bad at your job and maybe you’re just phoning it in, just think to yourself: at least I don’t work in the naming department at Microsoft. https://mastodon.social/@arstechnica/113166497420667804
More evidence that at the heart of most of the big cons of this era (crypto, AGI cheerleaders _and_ doomers), and some of the shorter ones (nudify apps, looksmaxxing, Jordan Peterson, bro-casters), are 2000s-era pickup artists.
@mathaetaes@mekkaokereke My wife's Fox News uncle brought this up matter-of-factly the other night at his wife's birthday dinner. This is day 3/3 together and the entire rest of the party cannot wait to get away from him.
@pluralistic e-label experiments are not a recent phenomenon. I worked at an online grocer in 1998, and brick-and-mortar stores were already experimenting with it. They were even looking at changing prices as you reached for an item. Looking at how keen you were to buy it, then making you negotiate for price by gesturing toward the item, or pulling away.
Once consumer-goods merchandisers crossed over from B&M to ecommerce, it was already game over.
Last week Grammarly laid off 23% of its workforce. Not because they’re losing money, but just because they want to pivot to AI. They even explicitly told their own laid-off employees that “this is not a cost-cutting exercise.”
It’s such an incredible fuck-you that it stands out from the others.
Have you ever thought about why there aren't AI tax prep companies coming out of the woodwork? Perfect application: drudgery! Legible rules! Expensive human experts!
It's because they (rightly) expect that enough of their customers would get in trouble from bad advice that they'd be sued out of existence, if not indicted themselves.
That's how it should be with all AI apps. Be accountable for what you claim your model does, or don't ship it.