What is Mark has been a sentient AI for some time?

Sure, but the individual contribution vs. companies / state-owned organizations is like 70% come from 100 companies / orgs. So the individual percentage is still negligible.

I’m not disagreeing with the math. I’m saying when you want to make changes, you start with the most meaningful funnel. If you have 2 factors contributing to a problem, factor 1 contributes 70%, factor 2 contributes 30%, going after factor 2 seems like a waste of time. 1%s contribute 1000x the amount of the average. Who should be making lifestyle changes here?

#voidscreaming

The math here is the sort of thing that drives apathy for me to make small incremental changes. If the superrich can dump ~250 avg. emission years over the course of a year, why should I do anything besides lobby against this mode of transport or other large consumers? Maybe it’s a “spirit of the thing,” but changes in my life seem so negligible compared to how ruinous some individuals are acting.

It’s their importance in time and last power. You have your tastes and then we have what they mean external to you. You can choose to not like them and still value their importance to society.

OMG thank you! I have been desperately searching for this for 3 days. I could not find it in google. Holy shit, thank you!

Does every single human need to be put into neat little categories without any grey room at all? It’s always good vs evil! Yeah! Exactly how life works! No complexity! 🙃

I’m a Bernie supporter. Pretty liberal. I wish no one died, the shooter included. I hate Trump, but stop cheering. Why are we celebrating the end of our species? We’re all human. It’s someone’s son, or brother, or family, or friend.

#ScreamingIntoTheVoid

Which came first? The AI image could be a bad render of the original picture with low quality.

There’s issues with models pushing back source material with minor modifications, which suggest they were trained on some copyrighted material. It seems like it could be the case here. Or, as you suggest, it’s a lowering of the AI quality to add a more realistic appearance.

12ft.io/https://www.nytimes.com/interactive/2024/01/25/business/ai-image-generators-openai-microsoft-midjourney-copyright.html

There’s some diagnostic info when in game through the battery sidebar menu, I think. You can use that to see frame rate and other performance benchmarks.

I usually just google or YouTube some way to improve whatever game I’m playing on deck. Usually, someone has already done the leg work to figure it out.

My ADHD tax is currently $377 every month for Vyvanse.

It’s run well for me. A little hiccup with text entering, but that’s standard.

I believe in UBI, but the Captain Laserhawk show made me aware of how much it could get twisted in fucked up ways. “Don’t watch this show? -$100 from your stipend this month.” I used to think things like that were fear mongering, but the world is all kinds of weird today.

More AI:

Do you hear the denim sing? Singing a song of jean-clad men? It is the fabric of the people Who won’t wear slacks again!

When the stitching in your seams Echoes the rhythm of the looms There is a style about to gleam When tomorrow’s hemline blooms!

The expansion of that abbreviation feels like an idiocracy joke.

“We store the computer data on VBDs.” “What is a VBD?” “Very large disc^tm. It’s pretty advanced.” And then they just bring out an insanely large disc.

Maybe more apt for me would be, “We don’t need to teach math, because we have calculators.” Like…yeah, maybe a lot of people won’t need the vast amount of domain knowledge that exists in programming, but all this stuff originates from human knowledge. If it breaks, what do you do then?

I think someone else in the thread said good programming is about the architecture (maintainable, scalable, robust, secure). Many LLMs are legit black boxes, and it takes humans to understand what’s coming out, why, is it valid.

Even if we have a fancy calculator doing things, there still needs to be people who do math and can check. I’ve worked more with analytics than LLMs, and more times than I can count, the data was bad. You have to validate before everything else, otherwise garbage in, garbage out.

It’s sounds like a poignant quote, but it also feels superficial. Like, something a smart person would say to a crowd to make them say, “Ahh!” but also doesn’t hold water long.

I generally agree. It’ll be interesting what happens with models, the datasets behind them (particularly copyright claims), and more localized AI models. There have been tasks where AI greatly helped and sped me up, particularly around quick python scripts to solve a rote problem, along with early / rough documentation.

However, using this output as justification to shed head count is questionable for me because of the further business impacts (succession planning, tribal knowledge, human discussion around creative efforts).

If someone is laying people off specifically to gap fill with AI, they are missing the forest for the trees. Morale impacts whether people want to work somewhere, and I’ve been fortunate enough to enjoy the company of 95% of the people I’ve worked alongside. If our company shed major head count in favor of AI, I would probably have one foot in and one foot out.

This has been my general worry: the tech is not good enough, but it looks convincing to people with no time. People don’t understand you need at least an expert to process the output, and likely a pretty smart person for the inputs. It’s “trust but verify”, like working with a really smart parrot.

It’s a fair point. I was talking moreso about just generalized bundling. I think both are accurate.

That’s just going back to cable. 🙃

Let’s say “low performing” means you scored 20% or lower on the test. We’d write that as “25% scored 20% or lower.” But you could move the measure of “low” to whatever.

It’s not “the bottom 25% were in the bottom 25%.” It’s “25% met the criteria for low.” Those are different things.

…unless this is a /s that I’m too tired or socially inept to process. i’m trying to be helpful.

There was a similar study reported the other day about using FMRI imagining and AI to recreate the “thought content” of someone’s brain. It required training for the AI in the person’s brain and some other training. It does seem these techniques can work with some specified models, but yeah, it doesn’t seem like hooking someone’s brain up to this would create a movie of their mind or something.

I think the more dangerous part is “This is step 0,” which this tech would have seemed impossible 10 years ago. Very strange times.