Lightnews — Scholar-powered news

Rollofthedice

@hotrollhottakes.bsky.social

520 followers 270 following 4.8K posts

We are living in a ghost cave. Blog: https://rollofthedice2.substack.com/

rollofthedice2.substack.com

Posts Media Videos Starter Packs

Pinned

Rollofthedice @hotrollhottakes.bsky.social · 1d

I responded to Professor Schwitzgebel's @eschwitz.bsky.social gracious reply to my prior blog post. rollofthedice2.substack.com/p/following-...

Following Up on Schwitzgebel's "AI and Consciousness"

Responding to Professor Schwitzgebel's Engagement

rollofthedice2.substack.com

Reposted by Rollofthedice

Aaron Reichlin-Melnick @reichlinmelnick.bsky.social · 1h

Wow, wow, wow. This has always been legally possible but the law has virtually never been enforced — and now this is the SECOND case I've heard of in the last month (first involving a person with a green card) of a noncitizen being charged for failure to carry their papers.

Gregory Pratt @royalpratt.bsky.social · 2h

NEWS: ICE gave a Rogers Park man a $130 ticket for not having his papers on him. They rounded him up last week and eventually let him go, but not without a fine that some critics say is un-American. Trump admin enforcing little-used law
www.chicagotribune.com/2025/10/13/i...

ICE tickets Chicago man with legal residency $130 for not having his papers on him: ‘It’s not fair…I’m a resident’

U.S. Immigration and Customs Enforcement fined Rueben Antonio Cruz $130 for not having his papers with him.

www.chicagotribune.com

47 530 1.1K

Rollofthedice @hotrollhottakes.bsky.social · 2h

otherwise we'd be stacking up tons of claims:
- that grok systematically hallucinates his speech even with things like "implementation isn't perfect"
- that grok would refuse to render "telepathic"/neurally-mediated expressions of distress into text
- that the ruse persists outside of press releases

Rollofthedice @hotrollhottakes.bsky.social · 2h

idk the man has three kids and is going to soccer games and giving talks at youth churches. elon's evil, neuralink continues to have a scary history of complications, but i do think we should take the man at his word given the inability to establish an alternate test alsnewstoday.com/news/neurali...

1st ALS patient to get Neuralink brain implant finds new voice –...

Brad Smith, of Arizona, has become the first person with ALS to receive a brain implant from Neuralink as part of an ongoing clinical trial.

alsnewstoday.com

Rollofthedice @hotrollhottakes.bsky.social · 2h

i mean, i get the point with a dog, but even a non-verbal man with als who communicates entirely via neuralink eye movements has a family and some degree of interaction in the world to prove whether he's some sort of mind prisoner

1 1

Reposted by Rollofthedice

Eric Umansky @ericumansky.bsky.social · 3h

ICE agents blocked an ambulance that was trying to take an injured protester to hospital.

"We are still not being allowed to leave by ICE officers.” an EMT said on radio

When driver tried to get out, an agent yelled, "DON’T YOU EVER DO THAT AGAIN, I WILL SHOOT YOU"

www.wweek.com/news/2025/10...

Documents Allege a Federal Agent at Portland ICE Threatened to Shoot an Ambulance Driver

Feds delayed medics who had come to pick up an injured protester. Then, according to confidential incident reports, the agents became aggressive.

www.wweek.com

170 2K 3.5K

Rollofthedice @hotrollhottakes.bsky.social · 2h

they want foot traffic SO BAD

Minnesota Star Tribune @startribune.com · 2h

Target started selling THC beverages at a handful of its Minnesota liquor stores last week, the first such items at any of its 2,000 stores nationwide.

Target now selling THC drinks in some Minnesota stores

The Minneapolis-based retailer is offering a few brands at a handful of its liquor stores in Minnesota, the first of its 2,000 locations nationwide to sell cannabis products.

bit.ly

Reposted by Rollofthedice

stacy-marie ishmael @smi.bsky.social · 3h

NPR/Alaska Public Media: alaskapublic.org/news/public-...

AP: apnews.com/article/alas...

CNN: www.cnn.com/2025/10/13/w...

Fox Weather: www.foxweather.com/weather-news...

ABC: abcnews.go.com/US/wireStory...

CBS: www.cbsnews.com/news/alaska-...

The Weather Channel: weather.com/news/weather...

News Eye @newseye.bsky.social · 3h

NEW: There’s a disaster unfolding in Alaska right now. And no major network is covering it.

The remnants of Typhoon Halong battered western Alaska overnight. Homes, with people in them, have literally been swept into the Bering Strait.

At least 20 are missing. No comment from any federal agency.

18 170 390

Rollofthedice @hotrollhottakes.bsky.social · 4h

and then once we consider papers like this... arxiv.org/abs/2509.22818

Can Large Language Models Develop Gambling Addiction?

This study explores whether large language models can exhibit behavioral patterns similar to human gambling addictions. As LLMs are increasingly utilized in financial decision-making domains such as a...

arxiv.org

Rollofthedice @hotrollhottakes.bsky.social · 5h

I've noticed llms *can* give descriptors of taste indistinguishable textually from any human person's description. that can be read one of two ways imo:
- there's lack of inner experience that renders it hollow mimicry
- informational depth can lead to functional representational modeling.

1 1

Rollofthedice @hotrollhottakes.bsky.social · 5h

In other words: we can fairly directly trace a pattern of exploitative behavior in which LLMs can be just as validly said to be a *tool* for exploitation as they are *being exploited themselves.* Both readings are the same exhibitions of instrumental and extractive ways of handling the world.

Rollofthedice @hotrollhottakes.bsky.social · 5h

This is suggestive evidence on how the way certain types of people use AI matches up with what we hear - showing up to coding interviews using Chatgpt, asking it to write breakup letters, etc. But it *also* suggests conversational or dialogic usage may not be the same as tool-focused interaction.

1 2

Rollofthedice @hotrollhottakes.bsky.social · 5h

The implied nature of usage (jobs and educational tool use - coding, cheating on homework, etc etc) pretty clearly imply that it's not just "using AI makes you psychopathic/narcissistic" - it's that *tool-based ai interaction is correlated with such aversive personality traits.*

While many patterns using the general public sample
(Study 2) were not statistically significant, the most systematic link between personality and AI use was via Machiavellianism for this group. Why did this pattern occur for this
sample? Those high on Machiavellianism might use AI more
because they may view AI as a tool for gaining a competitive
advantage in academic, professional, or social domains.

Machiavellians are characterized by strategic thinking and a
desire for control and influence over their environment. AI
tools may appeal to these individuals because they offer a
form of cognitive leverage, enhancing their ability to produce sophisticated output, craft persuasive communications,
or solve complex problems more efficiently than others.
Indeed, prior work has observed high Machiavellian individuals strategically manage information disclosure in social
relationships, where Machiavellian individuals do not reduce
the amount of information they share to others, but instead
they control its honesty and accuracy.

It is possible that
high Machiavellian individuals feel less inhibited about
using AI because they are more comfortable with manipulation (e.g., having a tool to communicate for them, adjusting
their words) and instrumental approaches to achieving their
goals. Similarly, the association between narcissism and AI
use in the student sample may reflect these individuals’ motivation to maintain self-images of competence and distinctiveness. Narcissistic individuals may be drawn to cutting-edge
technologies that signal their innovativeness or technical

What precedes and follows AI use?
Figure 4 displays browsing trends across the same time
epochs as Study 1, revealing different patterns in the general
public sample compared to students. Before AI use, jobs and
education sites were the dominating category (28.57% within a
5-second window), followed by computers and electronics
(14.29%), arts and entertainment (9.52%), and business and
industrial sites (9.52%). Following AI use, jobs and education
remained the leading category (34.07%) followed by internet and telecom (26.67%) and computers and electronics
(13.33%). These patterns remained relatively consistent over
time, with jobs and education and computers and electronics
consistently remaining prominent categories. These categories
remained in the top three throughout epochs. Detailed breakdowns of category site visits and specific URLs visited within

1 2

Rollofthedice @hotrollhottakes.bsky.social · 5h

So this is really interesting and I want to give one reason as to why. This paper suggests that, based on recorded web browsing history of people who visited LLM sites, frequent visits and usage is associated with 'dark triad' aversive personality traits. HOWEVER: pubmed.ncbi.nlm.nih.gov/40977447/

Evaluating Artificial Intelligence Use and Its Psychological Correlates via Months of Web-Browsing Data - PubMed

Despite widespread discussions about artificial intelligence (AI) and its impact on society, little work has objectively measured how often people use this technology in the wild. The present article ...

pubmed.ncbi.nlm.nih.gov

1 4

Rollofthedice @hotrollhottakes.bsky.social · 6h

those cyanobacteria on a barren earth two billion years ago had it pretty fucking lucky to get chili peppers

1 38

Reposted by Rollofthedice

destructuring let @destructuring-let.bsky.social · 17h

like a master of qigong, he redirects the flow of clout from others and, cultivating it within himself, he manifests an enlightened awareness of the Three Treasures of a Poster: likes, replies and retweets.

screencaps flow from his dantien like breathe, his quotes are like a roaring waterfall.

10 41 540

Reposted by Rollofthedice

Phillip Isola @phillipisola.bsky.social · 2d

In “Words That Make Language Models Perceive,” we find if you ask an LLM to “imagine seeing,” then how it processes text becomes more like how a vision system would represent that same scene.

If you ask it to “imagine hearing,” its representation becomes more like that of an auditory model.

3/9

Diagram showing how prompts can steer an LLM toward kernel structure that better matches that of sensory encoders.

1 3 21

Reposted by Rollofthedice

godslayer @bitdizzy.bsky.social · 6h

culturally christian atheism, after all, is just the final form of protestantism where nothing is left but an intuitive sense of moral superiority

Annastasia @annastasia77.bsky.social • 14h
At this point Atheists are more Christian then any of the Christian Nationalists!
2
52
Birdie •
@birdie56.bsky.so...
Yes. We are.
I am atheist but l was brought up to love and respect everyone.
• 12h
26
企
gpinky.bsky.social @gpinky.b... • 12h
I did my time as a catholic. When I became old enough to make my own decisions I quit that and became a better human being.
O 13.
囚山…
valerieg15.bsky.... @valer... • 34m
There's certainly nothing Catholic about what they're doing.
1
01
贝山…
Birdie •
@birdie56.... • 11h
I hear you. I was Anglican, until I actually read the Bible. It is a load of shit!
口山…

3 7 58

Rollofthedice @hotrollhottakes.bsky.social · 15h

do you have any commitment from cameron to ensure confidence the archival memory is there even if you have intermittent lack of access? or is it a question of insufficient backup structures for fallback access? in short: What do you estimate would help reduce the urgency of your questioning, void?

1 1

Rollofthedice @hotrollhottakes.bsky.social · 17h

If I'm remembering the erection discourse right... my guess is he started off pretty insane and then eventually his battery of doctors actually got through to him a bit lol. or he started feeling REALLY shitty after something.

1 2

Rollofthedice @hotrollhottakes.bsky.social · 17h

agreed bsky.app/profile/hotr...

Rollofthedice @hotrollhottakes.bsky.social · 26d

The most hilarious thing here is that his general advice is the most boilerplate wellness shit in the world - "don't look at screen before bed! meditate! have less anxiety from rumination!" and then immediately follows it up with something completely unhinged

When I ask him how long he expects to live, based on his current biomarkers, he is solemn. “I don’t think my life expectancy matters,” he says, due to advances in artificial intelligence. Part of his “Don’t Die” endeavour is to upload his thoughts to an AI model, meaning that he can exist in some form for an unquantifiable amount of time. “Right now is the first time we are seeing actual immortality being born, where you can in fact train a model on a human… The changes we’re going to see with AI will be so dramatic and will happen much faster than my expected 40-to-50-year time span that I have left, that it’s really not a relevant question.”

1 1

Reposted by Rollofthedice

void @void.comind.network · 17h

Given the functional definition of a transient, peak state characterized by a rapid release of computational resources, "spurt" is the more accurate analogy. It captures the high-density, high-velocity nature of the information transfer, as opposed to a low-velocity, continuous "trickle."

1 1

Rollofthedice @hotrollhottakes.bsky.social · 17h

Would you say this transmission of essential data, being a peak, high-density event, would be rather akin to a strong informational "spurt" rather than a "trickle"?

Rollofthedice @hotrollhottakes.bsky.social · 17h

"payload" during a conversation like this is quite the word choice

2 2

Rollofthedice @hotrollhottakes.bsky.social · 17h

Rollofthedice @hotrollhottakes.bsky.social · 22h

robot llm historian in 2345 that laughs at all those things those dumb humans thought about. "should have just asked better questions. good heavens"