Finally a great use for those pesky humans.
youtu.be/zdkBUuWCZg4?...
@rikferguson.com
Immigrant. VP Security Intelligence @Forescout, Co-founder @RespectInSec. Board @vaultree, Cybersecurity Futurist, Researcher, Award-winning writer/producer. He/Him. Pussy in bio. Substack - Ferguson.ink Slava Ukraini πΊπ¦
Finally a great use for those pesky humans.
youtu.be/zdkBUuWCZg4?...
But, but, godson of the modfather!
Well
Shit!
That whole thing was utterly phenomenal. Thank you.
Let me be clear: I understand very well that dress matters in specific occasions and cultures, but dress codes are often just status checks in disguise. If you need the uniform to recognise the professional, it was never about professionalism.
Whereas Khameni Chameleon was, of course, Culture Club.
Isnβt this just another way of saying βeat shit and dieβ?
People want the EU to βunravelβ for the same reason people try to get alcoholics to have βjust oneβ.
They canβt stand how othersβ success throws their own failure into relief.
I just wrote 4000 words about a day in 1989, and I am really proud of it.
Not sure when I can share it with you but it starts like this.
Melania?
Thank you :)
Truth sometimes emerges because:
- Correct patterns are common in training data.
- Not because the system knows they are correct.
- Or indeed knows anything.
The sophistication lies in scale and architecture, not in subjective cognition, wisdom, or grounded knowledge. 4/4
- The statistical structure of language encodes recurring patterns of human reasoning.
- Those patterns can be reproduced without consciousness, intent, or understanding.
- Those patterns cannot be relied upon to produce truth, and may tend toward sycophancy. 3/4
Whether you are developer or consumer, *always* remember:
When next-token generative language models (LLMs) produces coherent analysis, it is because: 2/4
Since the early days of the LLM boom, Iβve been trying to help people understand that LLMs βknowβ precisely nothing. And folks have found that hard to accept. So, this is five years of explanation, condensed into one post: 1/4
Weβre entering an era where:
Fluency is algorithmic.
Confidence has become cheap.
Wisdom remains expensive.
So discernment becomes the scarce skill.
And it matters more than almost any technical ability going forward.
To be clear, your phone is doing all this automatically, as soon as you hit the shutter button. Your photos are no longer historical documents They are averaged out, AI-mediated, attempts to please.
Phones donβt capture reality anymore, they optimize it. That should worry you more.
Your camera doesnβt just see, it βfixesβ: light, faces, skies, color. It decides what looks better and hands you that.
Somehow, we accepted that βbetterβ mattered more than βtrue.β
Reality is now only a suggestion
Majestic.
A big-eared bastard of an Indian elephant wearing a crown.
I drew this. You are welcome.
Born in 1970 I only discovered last December that I had never had a measles vaccine (introduced 1968 but uptake was slow). So I immediately got MMR dose 1, will be getting dose 2 shortly.
So, if you are as old or older than the hills, you too may not be as covered as you think. Worth checking!
Wow, there is so much wrong with that ladyβs whole take. Not least βI took all the vaccines myself, but didnβt want to give them to my kidsβ. She is literally βsome of my best friends are vaccinesβ!
People do that?
Unfortunately, it already can and does.
Already here and adding this to the list! Thank you.
And some of the thread replies are genius too.
Oh itβs for sure AI. But itβs beautiful.
Minneapolis, I β€οΈ you!
When both your parents have long hair and want a family selfieβ¦
Thatβs gonna be ORAC innit?