They pushed phone voice recognition out to their corporate cloud, but now they want to run AI on the phones 24/7?
That’s pretty sus.
They pushed phone voice recognition out to their corporate cloud, but now they want to run AI on the phones 24/7?
That’s pretty sus.
Also, is it feral?
I mean… shouldn’t they be thinking of ways to cut down on human caused mistakes? Like, give you guys better, less rigorous hours? Get more people in the field by helping people go to med school (financially, I mean)?
they really are coming for all of us, however skilled we might be. If they can find ways to pay ALL of us yes, and keep making $$$ for the 1%, the rest of us can fuck off…
What could possibly go wrong?
Really, though, “my boss wants me to fill two pages with something that looks plausible but it’s boring and I already know nobody is ever going to read it” is kind of the best use case for ChatGPT.
I mean, true enough… But I don’t think that the author is taking that kind of mindset into account? It’s more of a “see, we TOLD you haters” kind of tone, to my mind, at least. For them it’s just evidence that proves that this technology is disruptive and fast spreading, and it will transform how businesses operate, but they don’t really seem to care about the implications of that if true. Like, the underlying assumption is just that anything the changes things is by definition good? Which, for me, that aspect is the most frustrating part. There is just so much “charge ahead” attitude at play here, that I feel like we can’t even really have an actual conversation about whether any of this is actually helpful for humanity, or just another thing designed to move wealth upwards.
The comparison to spreadsheets is interesting? I’m not a historian of business history, so I can’t really speak to whether or not that’s apt, but I guess it could be?
We had a problem at the end of last year where people were using ChatGPT to write their own self-assesments for year-end reviews. The issue was that some were uploading company-proprietary or sensitive data to generate the outputs. IT had to block all access.
The Congress is such a great and underrated film. Robin Wright President Princess Buttercup is so awesome.
Not really. The boss might not read it (carefully) or may not be competent to revise it properly. Like when they specifically ask a person in their department who is competent to come up with, say a new safety directive or a company policy that makes sure that new regulations are adhered to.
Or the company doesn’t want to pay for somebody who is competent, so someone else is tasked and starts looking online for something that might fit the bill. And in good faith, too.
This sort of thing is happening right now, and there are disasters waiting to happen because of it.
If a new safety directive or regulatory policy falls under the category of things nobody is ever going to read, something has already gone terribly wrong.
I’m (first-hand) aware of a few cases where spreadsheets played a key role in multi-hundred million, if not billion, dollar losses at some firms. If the comparison is with badly applied software, with deeply opaque workings and the potential for massive liability then sure, I suppose the analogy works.
The article frames it as a industry transformative technology.
When your only tool is a hammer, all your problems start to look like nails…
I suspect there may be an interesting intersection between the Graeber’s concept of “Bullshit Jobs” and the use of LLM’s within those jobs. Couple that with Peter Turchin’s ideas about “overproduction of elites” and we might have a thesis.
I suspect that many of such “overproduced elites” are employed in Bullshit Jobs. If they and their underlings are box tickers or doing marginal, rarely critical, work to begin with, then we might see an industry transformed by their functional, if not actual, unemployment by AI. Now, will this contribute to further collapse of technical ability in organizations (Boeing) or will the competent side win? I suspect the former, given the ability of LLM’s to generate endless streams of plausible but land-mine strewn verbage and the patience of competent people in dealing with that.
Turchin suggests that:
“In order for stability to return, elite overproduction somehow needs to be taken care of – historically and typically by eliminating the surplus elites through massacre, imprisonment, emigration, or forced or voluntary downward social mobility.”
I wonder which item on that menu “AI” is going to best facilitate?
What you need to know about NYC congestion pricing.
New York City is set to implement congestion pricing in mid-June 2024, touting it as a way to unclog traffic-prone streets and improve public transportation infrastructure.
Make sure your car is tethered.
Via Cory Doctorow’s blog today…