@Alfred is it possible to embed inaudible commands into audiofiles to control AI systems like voice assistants.
@Alfred @Zevon or flash a visual for a fraction of a second...
https://arstechnica.com/cars/2019/06/spoofing-car-ai-with-projected-street-signs/
@Zevon I like this but am mystified at the suggestion that cognitive science was codified in 2003?! I have books from a decade earlier that reference books from decades before that. Guessing he means something specific. Anyway, onward.
@b4cks4w LOL I see that kind of revisionism fairly often in the field
@Zevon embodiment! my favorite! *listening* (with my _ears_) (which are part of my _body_)
@b4cks4w 🤖 haha!
@Zevon c. 1979m "Cognitive Science" p.44
@b4cks4w 👍🏼 rite! another example:
Dr. Margaret A. Boden OBE SCD FBA is Research Professor of Cognitive Science at Uni Sussex. She helped develop the world's first academic programme in cognitive science. She holds degrees in medical sciences, philosophy and psychology and integrates these disciplines w/ AI in her research.
I recommend her 2016 book
AI: Its nature and future
@Zevon On the virtual traffic jam he mentions, there are examples of AVs doing this already. I think it would get vicious if they were more prevalent. Good thing they won't be 😉 😈
https://www.bbc.com/news/technology-58928706
https://www.usatoday.com/story/tech/2024/08/15/waymo-driverless-cars-honking-parking-lot-video/74810195007/
@b4cks4w lol 😉 I mean unless intelligence or security needs to.
@Zevon Thing about AVs is the financial justification breaks down as they approach technical feasibility. LIDAR-based braking and lane keeping will/have reduced accidents by (a lot). So safety won't be the driver (ha ha). I don't know how many people need (or want in Amurika) their car to have a brain solely for convenience, and will be willing to pay the incremental and ongoing cost (assuming a subscription, because of course)
And is L5 technically feasible? Not alongside humans for sure.
@b4cks4w good points 👍🏼 I think the same argument could apply to robots in the home - i.e. robots built similar to the human form. Only the very rich will be able to afford them for the foreseeable future.
On another note: looks like the Secret Service have robot dogs simiilar to Spot patrolling Mar-a-lago.
@Zevon Humanoid robots make zero sense to me, so difficult, and for what.
My job includes robots so I'm slightly versed. A lot of the platforms are adding security monitoring because adding a mic and camera is trivial and the things can be operated 24x7. A floor mopper can move towards and record sounds that happen at odd hours.
The dog platform is essentially open source hardware and software (ROS) so I expect that to promulgate.
@b4cks4w 👍🏼 I agree.
@b4cks4w did you see the scene in Leave the World Behind, where US adversaries block highways and escape routes by hacking Teslas sending them crashing into each other forming blockades trapping citizens.
@Zevon No! But you wouldn't even have to hack them. Just toss crap on the road they can't understand. Which is almost anything. A pillow.
@b4cks4w A pillow!! 😱 Maybe Mike Lindell should be classified a national security threat LOL
@Zevon for sure! For so many reasons. That mustache
@b4cks4w it alone is a disqualifier 🤣 🤣 〰️
@b4cks4w I'm sorry, I am unable to display or link to specific external content such as images or articles. However, I can try to provide information or summaries based on the content you mentioned. Would you like a summary or more information on spoofing car AI with projected street signs as per the article you mentioned?