techt @ techt @lemmy.world

Posts

2

Comments

310

Joined

2 yr. ago

Weird number of downvotes here -- I thought they were meant for low-effort or non contributive comments, not an "I disagree" button. This person is giving a unique perspective as a subscriber (in this thread, anyway) and should be met with curiosity, I think. It is helpful to know that there are people who enjoy paying for it, so thanks for giving your opinion here.

I disagree because they have a dominant position for reasons other than having a good product -- they squash competition trying to make the space better while themselves actively making it worse. Subscribing means supporting that style of inhibiting innovation, not to mention the other user-hostile practices they embrace (extend, extinguish). They are an ad company and obligated to make a profit, I get that, but I refuse to abide this style of using investor money to operate at a loss for years while deceptively capturing the market before raising prices. If your product is good, it shouldn't need to be artificially propped up.

Does this mean any "both" situation is actually an either-or, because the choice is still between both or nothing? I don't like that

Marry a lady!

Also, pedantry: activate! It's not an either-or situation -- it's a both situation.

On continue.

Once you get to the second semicolon in the same sentence, I'm calling it journalistic malpractice.

8 commas

2 em dashes

2 semicolons

1 sentence

Finding someone beautiful or attractive doesn't challenge or reveal anything about your sexuality. You're free!

"I am sworn to carry your burdens"

Yeah I feel you. I don't think the content is necessarily bad, but LLM output posing as a factual post at a bare, bare minimum needs to also include the sources that the bot used to synthesize its response. And, ideally, a statement from the poster that they checked and verified against all of them. As it is now, no one except the author has any means of checking any of that; it could be entirely made up, and very likely is misleading. All I can say is it sounds good, I guess, but a vastly more helpful response would have been a simple link to a reputable source article.

Without looking up a list, it's gotta be Robert-François Damiens right?

You have yet to suggest or confirm otherwise, so my point stands that your original post is unhelpful and non-contributive

The issue is you didn't confirm anything the text prediction machine told you before posting it as a confirmation of someone else's point, and then slid into a victimized, self-righteous position when pushed back upon. One of the worst things about how we treat LLMs is comparing their output to humans -- they are not, figuratively or literally, the culmination of all human knowledge, and the only fault they have comparable to humans is a lack of checking the validity of its answers. In order to use an LLM responsibly, you have to already know the answer to what you're requesting a response to and be able to fact-check it. If you don't do that, then the way you use it is wrong. It's good for programming where correctness is a small set of rules, or discovering patterns where we are limited, but don't treat it like a source of knowledge when it constantly crosses its wires.

My prediction is they'll train an LLM on all his speeches and put before the supreme court that it can be president because it executes his will.

Gattacum

...drunk?

Yeah! Some of my favorite builds involve bows. However, I think suggesting Dark Souls to someone who hates melee and enjoyed Remnant is a trap. Bow gameplay in Souls games is highly janky and bow-only is primarily done as a challenge run. I'd never suggest that to a first-timer to the series, but that's just me :)

Don't fall for this one either

Bows are also viable.

Do not fall for this, OP

I didn't know this project was this far along, that's really exciting news for browsers and the chromium hegemony! Going to drop them some support when I can.

Chernobyl, 100%. I worked in the nuclear field and studied the real event and resulting consequences, and this series nails the details. They strike a great balance between explaining a complicated subject to the everyday person, exploring the political and existential tension, and providing an entertaining story narrative.

That's closer to what I was thinking. Most of the suggestions in here are outright vandalism which is an asymmetric response and not very defensible. But if you're physically capable, depending on traffic conditions, I'd feel like the safest route to go is over the top of the car. I'd do it for a car-shaped rock, so a car-shaped car is no different.