Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment Re:Reporters? (Score 1) 166

I've made this same point myself. When I was at Google I/O, I sat down for lunch with a table of Google engineers. AI came up, and the topic of jobs came up. I asked, "Let's say AI does actually replace the people that you think it can, after a few years, where does AI get the training data to keep up, or will we have a new wave of humans-replace-AI stories coming?"

To their credit, they confirmed that's a concern that no one seems to have figured out, and no one seems actively paying attention to simply because that's a "tomorrow problem".

In fact, decades ago Isaac Asimov actually foresaw a problem exactly like this in the story The Feeling of Power.

Comment Re:Translators and interpreters? Writers and autho (Score 1) 166

In most cases, the two sides can understand each other "enough" to complete the interaction without any AI. See also any grocery delivery or Uber driver in FL. The reason interpreters are needed is exactly because "most cases" don't require it, but when the case does require it, it's a really big deal when the wrong option is chosen. Interpreters even understand the subtly enough to know when to pause and ask a clarifying question. AI just powers through as if it knows it all.

Comment Re:Translators and interpreters? Writers and autho (Score 1) 166

Think interpreters especially. They're used in often politically charged scenarios between potentially unfriendly actors where a realtively simple misunderstand can have huge consequences. This quite literally happend between Couch F**ker and Zelenskyy a few months ago. Zelenskyy tried to speak English because it's well known Trump doesn't like interpreters, made a mistake, and Couch F**ker blew up on him in an entirely inappropriate, but consequential way. AI didn't even have to be involved for this to happen, but imagine the shit storm if it was.

Comment Re:Historians are not impacted by AI (Score 1) 166

LLMs are also not attached to specific point of view and do not care if they upset someone with their output

cough Grok nerfing cough ... The LLMs might not care, but the ones with the capacity to mindfuck them might...

Take what happening in Israel right now - no historian would touch it because of how controversial the subject is

Well, sure. But not for what you cite. Rather, it's not history yet as it's ongoing and the full details are still in flux.

Comment Re:Duh (Score 1) 73

I find bbb.org complaints to be pretty effective too. Google Pay once locked me out because I couldn't proved that I owned PNC Back... because I don't... I had an account at PNC. Seriously, this is what they sent me:

Thank you for taking the time to submit your information for verification. Unfortunately, what you submitted can’t be used to verify your account details and you will need to re-submit the document using the secure form here. Specifically, we require that you submit a certificate of incorporation for PNC Bank.

Spent almost two months trying to resolve it via their chatbot support and just constantly got told, "your issue has been submitted". Fired off a BBB complaint with all the details, the dates and times I tried to resolve it via Google's support, and the reference numbers for each time I contacted them. Less than two weeks later the BBB notified me that Google responded stating my issue was resolved and was asking for permission to close the ticket. They actually did fix my accout.

Comment Re:I've said it before... (Score 1) 20

I'm confused by the attacking nature of your reply. You said "don't complain, you brought..." Asking sincerely, at what point do you believe I was complaining rather than simply attempting to make a point?

Likewise, I agree many wear glasses (because they can't see properly otherwise), and many choose to wear sunglasses (to protect from bright light and some because they think it looks cool). I don't think either addresses, in any remote way, my point that not nearly enough have ever chosen to put glasses on their face for technology reasons to support a commercially viable venture.

Comment I've said it before... (Score 3, Interesting) 20

People will almost always pay more money to avoid having to put something on their face than they will willingly pay to put something on their face.

See also (not comprehensive by any stretch):
* 3D Movies
* 3D home TVs
* Quest / Metaverse
* Google Glass
* Apple Vision Pro
... vs...
* Contact Lenses
* Lasik

One group is quite popular and successful, the other is full of commercial flops...

Comment Re:Fentanyl - more deaths each year than Vietnam w (Score 1) 268

How many second hand fentanyl deaths have been reported? Because secondhand smoke is estimated at about 48k, or a bit under 70% of the total fentanyl deaths. Yet again, no global trade war. No upsetting of the economy. It seems like you have latched on to one single issue like a puppy who just fetched a ball and those around you need to tell you... "drop it... drop it... drooooooop iiiiiit...."

Comment Re:Fentanyl - more deaths each year than Vietnam w (Score 1) 268

Which part of my "fraudulent straw man" was wrong? That you are willing to support a global trade war? That you are willing to risk massive inflation? That you are willing to risk the damage to both our economy and our global standing? That the total number of deaths in which you cite is 15% of annual deaths by tobacco? Or that you are willing to do it in fidelity to an orange man?

Please, inform me. Which part of that was the so called "straw man" that was fraudulent?

Comment Re:Fentanyl - more deaths each year than Vietnam w (Score 1) 268

While you keep frothing on about fentanyl and about how it's a "crisis", there are an average of 480,000 deaths caused by cigarettes every year in the US. Just to be absolutely clear, you're willing to support a global trade war, a risk of massive inflation, and the damage both to our economy and our global standing because a talking orange convinced you that 15% of the number of cigarette deaths annually is that important?

Slashdot Top Deals

"Unibus timeout fatal trap program lost sorry" - An error message printed by DEC's RSTS operating system for the PDP-11

Working...