AI winter is well on its way (2018)

(blog.piekniewski.info)
12 points darkhorse13 | 7 comments | 05 Jul 25 18:28 UTC | HN request time: 1.467s | source | bottom
1. ctoth ◴[05 Jul 25 19:47 UTC] No.44475046[source]▶
The formula is so consistent: cherry-pick some disappointments, interpret any pivot as retreat, claim the hype is unsustainable, predict imminent collapse, position yourself as the sober realist. Rinse and repeat every 6-12 months, adjusting the timeline when the previous prediction fails to materialize.
replies(1): >>44475223 #
2. alganet ◴[05 Jul 25 20:17 UTC] No.44475223[source]▶
His prediction of self-driving cars being a fiasco does not need any adjustment, and it stands after 7 years.
replies(1): >>44475876 #
3. JLO64 ◴[05 Jul 25 21:54 UTC] No.44475876{3}[source]▶
Living in LA where there are plenty of Waymo cars constantly driving around, I strongly disagree with that.
replies(1): >>44476039 #
4. alganet ◴[05 Jul 25 22:17 UTC] No.44476039{4}[source]▶
Yes, 400 of them.

https://en.wikipedia.org/wiki/Waymo#Robotaxis

LA has more than 6,000,000 cars.

That is how strong you can disagree with me. 400 out of 6,000,000. And I'm being kind by considering just LA.

replies(1): >>44479475 #
5. joak ◴[06 Jul 25 10:33 UTC] No.44479475{5}[source]▶
What counts is not the total numbers of cars today but the growth rate of this number.

Ten doublings is 1000x

replies(1): >>44480813 #
6. alganet ◴[06 Jul 25 13:50 UTC] No.44480813{6}[source]▶
Waymo was founded in 2004. That's a very slow growth rate.

So far, the prediction made in 2018 stands. Self-driving car companies lose money, and are a fiasco.

Maybe that will change in the future, but that's the future. We'll talk about that then.