Lightnews — Scholar-powered news

Reposted by ahdok

Huffslove @huffslove.bsky.social · 22h

Silksong!

63 970 3.6K

Reposted by ahdok

🌱Sita Art🌱 Comms open @sitaart.bsky.social · 1d

The Prismatic Kobold Scholar

Illustration: a lovely scholar kobold with pale white and prismatic scales. She carries two arcane scrolls, a fancy tome and wears a colorful pair of glasses.

10 320 1.1K

ahdok @socksandpuppets.com · 16h

Alright, that's my ted talk. If you read it, thanks!

1

ahdok @socksandpuppets.com · 16h

It's the same argument with critical thinking as you made with doing art.

AI doesn't "make you a better artist" by doing the hard parts for you, and it doesn't "make you a better thinker" by doing the thinking for you, it just offers you an alternative to practicing and growing yourself.

1 2

ahdok @socksandpuppets.com · 16h

Microsoft put out a study months ago showing that people who relied on their copilot AI assistance tools showed a significant decline in critical thinking skills - because they stopped trying to figure stuff out themselves, they stopped practicing *thinking* and just let copilot do it for them.

1 2

ahdok @socksandpuppets.com · 16h

As for "supercharging critical thinking" - it's quite the opposite. You develop and maintain critical thinking skills by challenging your brain. Engaging with a problem and figuring it out yourself makes you better at figuring things out yourself. AI shortcuts that process to just give you an answer

1

ahdok @socksandpuppets.com · 16h

I think this is one of the biggest dangers with AI. People are just relying on what it says and trusting the outputs, because it *seems* knowledgeable, and they tend to do that when they don't have the skills to vet the outputs and check they make sense.

1 1

ahdok @socksandpuppets.com · 16h

I think all this buzz-speak around AI is just empty marketing. They're promising you that it's really amazing and intelligent, and they've built something that LOOKS like that right up until you actually know what the hell you're talking about, and then you realize it's a facade and a farce.

1 2

ahdok @socksandpuppets.com · 16h

The sum total of blending all this together isn't "superhuman intelligence" it's "average intelligence" - it doesn't have any way to distinguish the quality of the information it ingests, and it's weighted towards common responses, not accurate ones.

1 1 5

ahdok @socksandpuppets.com · 16h

It's basically mashing up all the "information" out there that exists that relates to your query, and spitting out a collage of the inputs, but those inputs include scientific papers that disagree with each other, wikipedia, reddit, the Onion, twitter shitposts, etc etc etc.

1 1

ahdok @socksandpuppets.com · 16h

If I ask ChatGPT about a topic I have no understanding of, it gives me some plausible looking response, but I lack the expertise to know if it's right.

Whenever I ask ChatGPT about a topic I took my degree in, I *always* see that it's responses are just... wrong? Like, there's so many flaws.

1 3

ahdok @socksandpuppets.com · 16h

I enjoy a lot of this, but I want to focus on the idea that AI is supercharging people's critical thinking skills

Sam Altman says that talking to OpenAI's new models is like talking to an expert with a PhD for every subject

But, for subjects where I actually have expertise, my experience differs.

1 4

ahdok @socksandpuppets.com · 17h

npr don't employ him any more, so I guess you got your wish?

1

ahdok @socksandpuppets.com · 1d

I find it really hard to figure out what makes empathetic sense for a souls-like protagonist to believe or think, because "if you die you just respawn" is often in-universe knowledge they have, and I have no idea how I'd handle learning that in my life...

1

ahdok @socksandpuppets.com · 1d

OpenAI (of all people) released a study last month saying they've proved that it's actually impossible to avoid hallucinations with this technology and the way they've designed it.

Wrong information is a baked in and inevitable part of the way it works.

So it's as useful as a magic 8-ball really.

1 1

ahdok @socksandpuppets.com · 1d

Well yes, it's just randomly compiling common responses to a string

The problem is it also regularly gives very bad advice (like eat rocks or put glue on your pizza), and has no capacity to distinguish

If you rely on it for societal stuff, it'll mostly just agree with you and reinforce your biases

1 1

Reposted by ahdok

David Revoy @davidrevoy.com · 1d

The Centaur and the Siren

#webcomic #krita #miniFantasyTheater

A comic strip in four panels:

Panel 1. On a shore, a beautiful mermaid and a young centaur meet. They look at each other.

Panel 2. Close-up portrait of the two protagonists, both look away, blushing.

Panel 3. Later, at sunset, the two admire the sunset together, side by side. They're a couple.

Panel 4. The happy couple watches their two half-human, half-seahorse children playing in the water.
Narrator: A few years later...

7 81 320

ahdok @socksandpuppets.com · 1d

And like, let's say the argument is "this will be hard to get done." and you don't think he can accomplish his whole platform...

The actual vote here is still a choice between someone who's actually going to try and make things better in new York, and Andrew Cuomo?

ahdok @socksandpuppets.com · 1d

Right, but also he's explained how his plan will work in his platform and in videos many times. His part of this is to secure the funding for the program, then work with the MTA to implement.

Since he managed to do this as a pilot with the limited power of a state assemblyman, we know he can do it.

1

ahdok @socksandpuppets.com · 1d

It won't make everything perfect overnight, but it will make things a little better. The reason I respect him is that the things he's promising are sensible, and actually achievable

They're things within the remit and power of the mayor. Doing them is just a matter of picking a mayor who wants to

1

ahdok @socksandpuppets.com · 1d

One of the reasons I like this guy is that he's not doing the Trump thing of promising the moon on a stick. "Everything will be better under me." with no solid proposals.

Mamdani's platform says "we're going to do this specific thing, here's the reasoning why, here's how we'll pay for it."

1

ahdok @socksandpuppets.com · 1d

Sure. It's just an example of how ridership affects wait times, that's all.

Mamdani's platform here is all stuff under the Mayor's control, and he explicitly states that in one of the videos here. One of the results of his policies here would be an increase in ridership, which would help wait times

1

ahdok @socksandpuppets.com · 1d

This is explained and allocated in the platform. No tax increases for regular New Yorkers. I think he said it's paid for with a small increase to the business tax rate for large businesses.

3

ahdok @socksandpuppets.com · 1d

And, of course, because ridership is high, car traffic along that road has dramatically decreased too. Before it was two lanes each way, and constantly at a standstill with the amount of cars. Now it's one lane (for cars) each way, and traffic is so light you can just cross on foot anywhere you want

ahdok @socksandpuppets.com · 1d

A couple years after I moved, the council restricted car traffic in the city center, and they painted dedicated bus lanes on all the roads along the route. Busses became the easiest and fastest way to get into town.

Ridership skyrocketed, and now there's a bus every 10 minutes rather than every 30.

2