I've been going through these videos for some work reasons, and they're so well done. Just great work all around by @astrokatie.com and @perimeterinstitute.ca.
05.03.2026 21:17 β π 92 π 18 π¬ 3 π 0I've been going through these videos for some work reasons, and they're so well done. Just great work all around by @astrokatie.com and @perimeterinstitute.ca.
05.03.2026 21:17 β π 92 π 18 π¬ 3 π 0Because... with this crowd, one never knows.
05.03.2026 21:36 β π 1 π 0 π¬ 1 π 0
Oh, the car.
Horsies (Equus caballus) are a whole other thing. Maybe I misunderstood what she was spending money on.
$60k is a high end Mustang now...
Only the best for this crowd.
OK, so Noem is out.
But the person behind the policy to begin with (Miller) is still employed in government and is pushing the toxic DHS agenda that Noem carried out.
LOL!
To be completely honest, I've often wondered about the feasibility of making a clone in metal with modern underpinnings.
A theorem...
Those theorems are based upon assumptions that the right network topology exists somewhere within the matrix, even if it is undefined a priori. You are also assuming that the right network topology can be found, and there is no evidence those topologies match neural ones.
I donβt think itβll affect the GOP Senate majority. The Oklahoma governor appoints an immediate replacement if Mullin resigns. (Assuming Mullin is confirmed.)
05.03.2026 19:24 β π 1 π 1 π¬ 1 π 0I think Oklohoma has a near instant replacement rule (Unfortunately), and it must be a member of the same party.
05.03.2026 19:22 β π 1 π 1 π¬ 0 π 0Heh heh hehβ¦
05.03.2026 19:33 β π 1 π 0 π¬ 1 π 0Letβs mos def chat.
05.03.2026 19:33 β π 0 π 0 π¬ 0 π 0
Mullin is replacing Noem?
Again, the thing that offends me the most is the gross incompetence of everyone.
Also, this is going to shrink the GOP supermajority to razor thin margins.
Going into the midterms this is good.
Entrance of the Grand Egyptian Museum, with a 'hanging' obelisk front and centre.
NEW The Grand Egyptian Museum displays c.50,000 ancient Egyptian artefacts from across its entire history, with major exhibitions Tutankhamun's life, death and identity.
Our latest #ReviewArticle from Martin Odler explores the "Truly Grand Egyptian Museum".
π doi.org/10.15184/aqy...
πΊ#Archaeology
OMG
05.03.2026 18:54 β π 1 π 0 π¬ 1 π 0
More than 1 in 5 new car buyers are taking out loans of 84 months or longer... on @roadandtrack.bsky.social
www.roadandtrack.com/news/a706218...
Oh, I am with you on good people being deceived. I don't think they are evil. Not by any stretch.
05.03.2026 18:37 β π 0 π 0 π¬ 0 π 0As someone who works in complex biological neural networks, I'll disagree.
05.03.2026 18:36 β π 0 π 0 π¬ 1 π 0Neural network (biology) != Neural network (machine learning)
05.03.2026 18:00 β π 1 π 0 π¬ 1 π 0
LLMs do not reason. LLMs do their work through linear and matrix algebra and statistical weighting.
That is not how biological neural systems work.
There are so many people who have been completely bamboozled by the anthropomorphic outcomes of LLMs, even through the naming of the tools/interface.
They are not neural nets like our brains.
They are not neural nets like our brains.
They are not neural nets like our brains.
They are not neural nets like our brains.
They don't "reason". LLMs are not built the same as brains.
I will vehemently disagree with you on that point.
Huib Petersen, an artist who lives on this block, is the artist responsible for these awesome crocheted creatures. www.petersenarts.com
05.03.2026 17:45 β π 9 π 5 π¬ 0 π 0
That is orthogonal to my point.
Yes, you had a successful outcome in a relatively simple system that outperformed the mechanic working on your car.
That does not mean that we should be basing targeting data that will take human lives in the outputs of under informed models that cannot reason.
Only recently are we building in confidence intervals into LLM based decision machines. And even then, they are making wild guesses with under informed training data.
05.03.2026 17:41 β π 1 π 0 π¬ 0 π 0
That is a relatively closed system with a known set of variables.
More complex systems (biology, human behavior, unmapped networks, or any system with unknown edge-node graphs) introduce combinatorial levels of variability in potential solutions.
If they make poor decisions, I would not trust them with my well being.
If I would not trust them with my well being, why would I trust them with larger infrastructures that impact others well being?
Certainly. But you were on Dunn & Bradstreet and got your contract through a transparent bid process, I would hope.
05.03.2026 17:17 β π 1 π 0 π¬ 1 π 0Logic from a classical electrical engineering perspective, not from an intelligence analytical perspective perhaps. Intelligence and targeting can certainly be informed by LLMs, but it is a misuse of the technology to ask them to give a logical explanation for why or why not actions should be taken.
05.03.2026 17:14 β π 1 π 0 π¬ 1 π 0This is where that $143M contract went to...
05.03.2026 17:08 β π 15 π 5 π¬ 2 π 0
You CAN query them to see the thread. You CANNOT query their logic, despite attempts to anthropomorphize tools like "Chain-of-Thought" or "Hybrid Reasoning".
The failure here is in the interface, the application, and the implementation.
Y'all make poor decisions.
The people in this administration have a demonstrated track record of abusing the animals around them, the people around them, the institutions around them, the people in those institutions, and the people dependent upon those institutions.
Why should ANYONE trust them?