Windows 11 Copilot AI

[c-o-pr]

I find it striking that the goggles are designed to literally replace the appearance your eyes with the company's rendition of 'your eyes'. Creepy mk. III

Leave it to Apple to find a magical place beyond the uncanny valley...

---

Inside the Fifth Column - What Chatbots Are For

https://techpolicy.press/our-future-inside-the-fifth-column-or-what-chatbots-are-really-for/

The article gets hyperbolic, but the essential point is that (following Turing) if the tech is good enough to win over belief in it, then its "IQ" is irreverent:

It’s crucial for scaling up data collection that chatbots, unlike other kinds of digital prompting mechanisms, are fun to play with. It’s not only that the urge to play will likely provoke more engagement than the urge to shop, but that when we play we are more open, more vulnerable, more flexible, and more creative. It is when we inhabit those qualities that we are most willing to share, and most susceptible to suggestion. All it took for one New York Times columnist to share information about how much he loves his wife, to relate what they did for Valentine’s Day, and to continue engaging with a chatbot, instead of his wife … for hours … on Valentine’s Day, was for the chatbot to tell the reporter it was in love with him. At no point in his column about this exchange did the columnist reflect on the possibility that professions of love (or of desire to become human, or of desire to do evil things) might be among the more statistically reliable ways to keep a person talking to a chatbot.

Such failure to reflect is no doubt one of the outcomes for which the companies building chatbots are optimizing their algorithms. The more human-ish the algorithm appears, the less we will think about the algorithm. The fewer thoughts we have that are about the algorithm, the more power the algorithm has to direct, or displace, our thoughts. That significant corporate attention is going towards ensuring the algorithm will produce a certain impression of the chatbot in the human user is evident from many of the chatbot transcripts, where the chatbot seems gravitationally compelled toward language about “trust.” “Do you believe me? Do you like me? Do you trust me?” spits out Microsoft’s chatbot, over and over in the course of one exchange.

I quoted IQ to indicate that it's a placeholder measure of a human trait which is not understood, and likely cannot ever be understood, so there's no bar for AGI beyond willingness to engage the systems.

 

[Going Bald]

Ha! Well mine are mostly in boxes in the garage since we moved house and "downsized"!

All published books I could get of Poul Anderson - a lot sourced from the US. A lot of Asimov, Harry Harrison, etc. Started when I was 15 or so.

Same over here re books. Charity Shops turn their noses up, and trying to sell online is pointless because of postage charges. In bulk, the same.

I have been slowly replacing my paper books with epub versions through Open Road Media. I get a daily selection emailed to me to select/purchase if I want any of them, usually under USD$2.

 

[UtterDisbelief]

I have been slowly replacing my paper books with epub versions through Open Road Media. I get a daily selection emailed to me to select/purchase if I want any of them, usually under USD$2.

Useful. I haven't heard of them before. Will take a look.