@mirrikybird
What's the saying? I'm not afraid of the robots becoming sentient, I'm worried about who controls the robots now.
@Freelix2000
We were too afraid of Terminator-like AI and not afraid enough of AI-powered scammers, crappy mobile AI apps, dead internet, and business stakeholders asking you to use AI to do things that can easily be done without it
@the_schmoopsie
I'm just tired of AI "customer support" chatbots that only know how to direct me back to the FAQ page that doesn't have an answer to my problem so I have to navigate the horribly made webpages that keep directing me back to the AI chatbot when talking to a real human being is all I need to fix my problem within a minute
@tjoloi
What they really should be worried about is your AC suddenly blasting hot air because it's only been tested in cold exterior temperatures and the entire machine is just a fan.
@JonBrase
This is exactly the video that a post-singularity AI would make to keep us docile! 😂
@dieucondorimperial2509
Lovecraft was deeply afraid of air conditioners, he wrote a horror story about it.
@marccox8977
Definitely not too worried about using LLM for Chatbots, but we should all be worried about autonomous weapons, or WMD without a person in the loop because those hallucinations and iffy training data could make us all have a really bad day.
@GSBarlev
This analogy is actually really good, in that my Nest thermostat burned a ton of energy over the weekend having misinterpreted a natural language ask to "make it a little colder" as "drop the temperature by 10 degrees," and it took me too long to notice and check in on what was going on.
@BlommaBaumbart
I'm not too afraid about sentience, but an actor (even if it's just some stupid program without anything close to intelligence) which can propagate through the IoT and adapt to new conditions, while it cannot be stopped by means of physical restraint, is still a novel threat.
@tyemich8820
Why is everybody concerned with AI outsmarting the corporations instead of them using the AI to create a distopian world
@M_Alexander
If AI was smart enough to take over the world I might be less concerned that people are already putting it in charge of things
@Atoll-ok1zm
The biggest threat AI poses is being used by idiots where determinism is vital. Like you can do it, you can use ML stuff in applications like avionics for say, computer vision. But it's difficult to be able to prove it's sufficiently reliable.
@NameRealperson
Gurg was right, we should never have split the rock.
@jon9103
The analogy doesn't work, AC is well understood by some people, AI on the other hand is complex in ways where we can't fully predict its behavior. It doesn't have to outsmart us to be dangerous, it just has to be allowed to control potentially dangerous systems and then behave in an unexpected way. There is no need for malice when stupidity can be even more dangerous.
@TheTrevorist
The way the "office lady" decided that an air conditioner is the equivalent of evil ai. Just take your iron pills Miranda 😅
@mymyscellany
I mean... kind of. I agree there are many people who have deeply unrealistic ideas about what ai is or what it is capable of. But I do think AI safety is a big deal and that it is LIKELY in the future we will lose control of many AI agents. A big thing here is timescale. I'm not worried about AI now, I'm worried about AI in 40 years.
@kicmorikic0
To be fair the AI safety is a fascinating topic to think about and fun to speculate on 😊
@bassman9261995
Ac company: “ok so we need our programmers to put a crypto trading AI in our thermostats to keep up with technology trends”
@NinjaLobsterStudios
"Oh, but hey did you ever hear this story about a computer and a lightning bolt?"
@albertatech