Forgot your password?
typodupeerror

Comment Re:Yes, we should be concerned about these things (Score 1) 108

*autonomous* AI operation of things like weapons

But when you are sitting in your F-35 and you receive targeting orders with some coordinates, how you you know what system or person was behind that decision?

And just like "autonomous cars are safer", AI based target selection might be safer (and less biased) than humans make. One can make the argument that an AI system could have integrated data from more sources and not called for the bombing of the Chinese embassy in Belgrade.

Comment Re:Yes, we should be concerned about these things (Score 1) 108

How do you know if it's AI? Are you going to come in and inspect my workplace? Audit my (company proprietary) tools source code?

Every time I read a post about how good and efficient AI is, and we all need to adopt it, I wonder about motives. If I had some neat toolset that enabled me to outproduce my competition, I'd keep my mouth shut. 'Sure. Give me the requirements and I'll run back to the shop and code something.' And it turns out that my code is more efficient and less buggy than everyone elses'. How do I do it? I guess I'm just good. So hire me.

All too often, when someone promotes a neat app that they've written, using AI, flawlessly, in a few minutes, it turns out that they work for an outfit that produces AI tools (or middle-ware). Or provides consulting services to adapt AI to existing workplaces.

The top level AI companies (Google, Anthropic, OpenAI) are basically doing their best to shove AI down the production pipelines. And get the lower tier app producers to load up on it and consume lots of tokens. In much the same way that auto manufacturers pushed product out onto dealers' lots.

Comment Blind (Score 1) 208

Sounds interesting. A series of social media boards, oriented around each companies culture. Either private, for inside discussions. Or open to the public. So they can learn a bit about the organization.

I used to work for an outfit that had an internal Usenet server. With some private discussion groups (restricted to employees) in addition to the worldwide Usenet (somewhat filtered of the NSFW stuff). It was like a giant rumor network around a virtual company water cooler. Some managers hated it. But many appreciated the value of being able to lurk and monitor employee sentiments.

Sounds good, until I read this:

Blind is capitalizing on the increased interest with new products. It plans to unveil a service called Blind AI, which will allow employers to simulate their workers' reactions to certain changes, like a stricter in-office mandate.

So, management can astroturf their proposed policy changes and try to guide employee opinions? Back in our Usenet days, management sycophants were rather easy to sniff out. And ignore. And most of the managers were smart enough to leave the system alone, lest they pollute the unbiased nature of the opinions posted.

Comment Re:Let's Just Let It Happen, Perhaps? (Score 1) 146

people who don't trust vaccines

Generally, these people tend not to trust other sorts of drugs either. The exception was the old left wing "anti-vaxers" who were following some coldly calculated logic: Vaccines may work. But there is also a very small probability of side effects. Better that the less advantaged take a vax and build up a herd immunity which will benefit me and my offspring. And let those riff-raff contend with the autism and other side effects. My little precious is too valuable to risk.

Comment Re:Deliberate unrecoverable damage (Score 1) 146

people should have bodily autonomy when it causes no actual harm to others does not permit me to condemn others for doing stuff to their own body for their own reasons when it does no actual harm to others even if it's a context I don't agree with.

Unless we are expected to pay for the rehabilitation and ongoing care for those people who exercised that autonomy under some form of socialized medical care.

Comment Re:Unsurprising, To Me. (Score 1) 20

Right. A while back, CloudFlare was doing its "are you human" tests*. And then failing mid-test when it couldn't pull some stuff from an untrusted site (something like error-report.com, IIRC). Reason for the failure? That site had been identified by my (very consientious) ISP as a malware source and blocked.

*Actually, more of an "are you runing JavaScript, so our customers can upload some annoying tracking crap to your machine" test.

Slashdot Top Deals

"In the face of entropy and nothingness, you kind of have to pretend it's not there if you want to keep writing good code." -- Karl Lehenbauer

Working...