Can’t see anything I disagree with in her statements. Imo she’s spot on. The near future looks even bleaker now.
People using it to control their computer is dumb as fuck. I dunno why AI companies are pushing the idea when they already have a killer product.
The killer product being the lying machine or the deleting working code machine? I think there’s a small number of people for who these tools really fit into their workflows but they are not universal so there’s limited growth and they’re already wildly unprofitable.
Something like half a billion people use ChatGPT and competitors every week. It’s really hard to deny that these companies are successful. Doesn’t matter what the quality is, it’s clearly good enough that people continue to use it.
“Successful” would include “profitable”.
lol, have you seen public companies nowadays? Amazon took 9 years to turn a profit.
Because “user-friendly” UIs have successfully, market-wise, killed normal computing (like under Windows 2000, or even like “advanced users” under Unix-likes do, nothing complex or hard, not even harder than the “user-friendly” way, but very scary when you’re conditioned to think it’s not normal to edit configs or run commands ; it’s very stupid, one would think editing files or entering a few words and pressing “enter” are not godlike powers).
That had the (subjectively) positive results of enshittification and monopolized Web.
Replacing the “user-friendly” UIs with mobile-like UIs mostly failed cause those are simply inferior.
But agentic AIs seem the way to go so that the typical user would never ever try to form preferences of how they use things, their own habits and processes.
And yes, the bigger the heap, the easier to hide a microphone there, and each such level of obscuring and generalizing control makes the heap order of magnitude bigger.
Rich people have staff that knows all their secrets; assistants, bodyguards, maids, drivers, … Their staff might abuse their position to steal, blackmail, or even sign book deals. They still have the staff.
When AI agents become useful for normal people, normal people will use them.
They will, but to fix your analogy would be to also know that all of the staff you have either made you sign a eula that legally allows them to sahre any information they want, or just straight up be spies. If this was the case, I don’t think they would have that staff.
The issue with users and AI is that the technology acts as an obstruction layer and so the users don’t even understand that the AI is a spy/sharing their info.
Sadly, they will still use them.
“And if we’re talking about a sufficiently powerful … AI model that’s powering that, there’s no way that’s happening on device,” she continued. “That’s almost certainly being sent to a cloud server where it’s being processed and sent back. So there’s a profound issue with security and privacy that is haunting this hype around agents, and that is ultimately threatening to break the blood-brain barrier between the application layer and the OS layer by conjoining all of these separate services [and] muddying their data,”
irrelevant banter
Anyone else simping for this absolutely gorgeous CEO? Like damn.
How the fuck is this relevant?
She has amazing hair too.
She had me at “Meredith.” Everything else just solidifies my infatuation.
💕
Be still my heart.
