mediaeater.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

mediaeater.bsky.social

Dispute The Text

120 posts 311 followers 118 following

Posts 23 Comments 27

This of course, is a lie. It does not "understand" and will do nothing. This is what deception looks like.

submitted 2 days ago • 0 comments

iOS intelligence learning from wamo behavior. This is the actual ‘context window’

submitted 3 days ago • 0 comments

Just started receiving calls mid flight. In airplane mode. That’s a first for me FaceTimes too

submitted 4 days ago • 0 comments

www.microsoft.com/en-us/resear... Found that while AI can improve efficiency, it also reduces critical thinking effort and can lead to over-reliance and diminished skills. Suggests that higher confidence in AI is associated with less critical thinking, while higher self-confidence leads to more.

submitted 6 days ago • 0 comments

Adams must be removed. The city cannot sustain being governed for nearly a year by a Mayor who is being coerced by Trump admin in order to escape charges. This corruption poses a real threat to the people of the city. He should have resigned a while ago, but will not. So it’s time for him to go.

submitted 8 days ago • 1781 comments

it's the consistency and reliability that I like in openAI products

submitted 9 days ago • 1 comment

This mural is about to be wiped out.... please share www.10014.com

submitted 9 days ago • 1 comment

Valentino. Vertigineux. Breathtaking indeed, both couture and presentation, information integration as design esthetic. Swoon.

submitted 9 days ago • 1 comment

been testing chatgpt tasks -its been ten days of deceptive promises and daily emails only to be trolled each day - the is the fresh hell we deserve.

submitted 9 days ago • 0 comments

@marklevinenyc.bsky.social We could use your help with this

submitted 9 days ago • 0 comments

www.youtube.com/watch?v=3fTp... AI and creativity . 1:45 mark Pharrell talks nonsense

submitted 11 days ago • 0 comments

ghuneim.com/2025/02/10/t...

submitted 12 days ago • 0 comments

fever dream masquerading as prophecy: one convenient data point magically turns into a 10x annual cost reduction, while "super-exponential" progress gets promised with the same certainty as a late-night infomercial. Silicon Valley's church of perpetual growth finds its new gospel.

submitted 12 days ago • 0 comments

DeepSeek’s R1 reportedly ‘more vulnerable’ to jailbreaking than other AI models Narrator: This is a feature. First you short the market, then you destroy the democracy by giving foreign intel services seeking to influence us elections a wide open AI. Justice department said fine by them.

submitted 12 days ago • 0 comments

idiallo.com/blog/listen-...

submitted 16 days ago • 0 comments

Optical frequency combs aka rulers for light.

submitted 17 days ago • 0 comments

2600 Hacker Quarterly cover of a decaying Musk

submitted 17 days ago • 0 comments

www.ft.com/content/cf11... focusing on model abstraction vs deception. This is akin to consumers paying fraud protection because companies deploy products that are not trusted.

submitted 18 days ago • 1 comment

submitted 20 days ago • 0 comments

copyright.gov/ai/Copyright...

submitted 23 days ago • 0 comments

Seriously who is measuring llm models ability to deceive? They actively practice it. Deception Scores are needed now. chatgpt.com/share/679a94...

submitted 23 days ago • 1 comment

www.wipo.int/web-publicat...

submitted 24 days ago • 0 comments

They single most important AI benchmark that is nowhere to be found is understanding the models propensity for deception I have been working on basic methodology and results. Confidence is low that we are paying 'attention' <-- to what matters.

submitted 24 days ago • 0 comments