-
I guess you could say that Intel is keeping an “i” out for us.
1 user thanked author for this post.
-
I’ve been using WinCompose (https://github.com/samhocevar/wincompose), which has many intuitive features, both portable and installer modes, and an assignable compose key (default is the right <Alt> key). However, this looks like it might be easier to use, especially as I am familiar with AutoHotKey — thanks, Brian!
1 user thanked author for this post.
-
-
Thank you, Max Stul Oppenheimer for another stimulating article. As he states at the end:
… Imposing liability for damaging falsehoods might be the incentive required to build more reliable AIEs.
Sometimes responsibility is a good thing.Along similar lines, lawyers for public (and various) figures are claiming that any evidence against their clients could have been faked through AI. NPR has a good article about this trend. Elon Musk & Tesla are being sued by the family of a man who died while using the self-driving feature of his Tesla. Video evidence shows Musk saying that autonomous self-driving is safer than driving the car yourself. According to the judge in Huang v. Tesla Inc., 19CV346663:
“What Tesla is contending is deeply troubling to the Court…Their position is that because Mr. Musk is famous and might be more of a target for deep fakes, his public statements are immune… In other words, Mr. Musk, and others in his position, can simply say whatever they like in the public domain, then hide behind the potential for their recorded statements being a deep fake to avoid taking ownership of what they did actually say and do.”
This also means that anyone who can afford the lawyers, can use the deepfake defense to drag out proceedings and make it impossible for a not-rich individual to win a lawsuit, since trying to get to the bottom of such garbage will be prohibitively costly. Perhaps the answer here, too, is to impose liabilities.
What a virtual can of virtual worms is this thing called AI! This seems to be true for all its aspects. For example, medical companies say brave new medicines can be designed more easily with AI. But, so can biological and chemical weapons… and on and on.
-
-
-
Here is a different perspective on responsibility & morals as it pertains to AI, and how it will probably be affecting us in the very near term:
https://www.schneier.com/blog/archives/2023/04/llms-and-phishing.html -
The FTC commissioners are not unanimous in their support for the proposal. A dissenting statement argues that the proposal goes too far… The dissent quarrels, not with the seriousness of those problems, but with the scope of the FTC’s authority. In Commissioner Christine Wilson’s dissenting view, the transaction’s terms extend only to the terms by which the transaction is entered and by which it may be terminated — all else is “marketing.”
Not very surprising, given that Christine Wilson is an anti-regulation Republican appointed to the FTC by Donald Trump. She believes that competition, not regulation, is the best protection for consumers. Hah! Who is she kidding? The current situation, where consumers are tricked into subscribing to a service, publication, or product, and are then unable to unsubscribe without paying penalties in time and/or money, is a direct result of competition without regulation. As Max Stul Oppenheimer notes, the FTC proposal is very limited and covers only negative options marketing, yet represents a “big step forward.”
Some good news: Christine Wilson’s term was to run to 2025, but she submitted a letter of resignation on March 2, and quit the FTC on March 31, 2023 to protest against the actions of FTC chairperson Lina Khan and the Biden administration.
So perhaps the proposal has a chance. I feel it doesn’t go far enough. If some feel that the FTC should not have powers to prosecute corporations, I suggest they can use a carrot instead of a stick. They should reward companies that practice transparency, by creating a “Certified by the FTC” sticker or category that companies can put on their products or services.
2 users thanked author for this post.
-
-
-
According to OpenAI’s report, GPT-4 did better than 90% of human test-takers on the US Uniform Bar Exam. It outscored 93% of them on the college-admission test known as SAT Evidence-Based Reading & Writing (EBRW).
I’m not surprised. I have worked in a university setting in one job or another for over 40 years, and conclude that the writing skills of college students have significantly deteriorated. I notice that current students are used to talking to their devices, not writing on them. When searching for information to learn skills, most prefer to view a video rather than read text (of course, this might be the wise choice, since most people seem unable to write useful instructions). Is it so unexpected that a machine has better writing skills? Research shows that writing and speaking are supported by quasi-independent parts of the brain. To use a simple metaphor, the writing “muscle” for most people needs to be exercised more.
2 users thanked author for this post.
-
-
Not only cart-before-the-horse, but also a security issue. I don’t like having a network connection open before I can set up my firewall and anti-malware software!
3 users thanked author for this post.
-
I wrote an article abut three methods of killing Clippy back in 1999; it was described as “There’s no better homicidal how-to on the Web.”
That reminds of one of my favorite “Wait, wait, don’t tell me” radio episodes, the famous “Clippy must die” exchange, where Clippy is taken for a ride in a car:
https://www.youtube.com/watch?v=NxcmoLKVd60
(the part you want starts about 2 minutes into the youtube audio)1 user thanked author for this post.
-
Author
