The Itsukushima summit to town is like 30-40 min walk on a solid paved ground

Heh. You are right. https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fmaps.app.goo.gl%2Ftevw9p... shows some of the easy path, and in the early evening it would also be a very lovely (and quite short) stroll! A situation where your GPS leading you typically astray became a welcome event instead.

Without taking away from your valid point:- I think the Aussie/UK slang "poor little diddums" fits her more than 'karen'. Karen is that well-known and quite typically horrible, screamy, middle aged woman usually demanding to talk to the manager as her face turns redder and redder, whilst diddums is more "dawww, you poor treasure, that must have been soul-shattering for you" :p

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fpatents.google.com%2Fpat... It's above my pay grade, but seems to be a UI patent? A system designed to render a collapsable tree on a server and present it to the user's computing device which might be VR with Haptics.

From the paper:

WAN2.1-T2V-14B consumes over 415 Wh

So, generate slop with this model for a whole hour and it uses 415 watts of power? What they don't seem to mention - that I could find - is how many minutes of slop they generated in that hour of use.

To me, a pure Watts/Slop metric would be more useful, and far more compelling. To that end, I propose the Zuck as a measurement to encapsulate watts/slop-second. If a 10 second video takes 10 real minutes at 415Wh then that's 69 watts or 6.9 Zucks of power (and water).

When I go to the doctor's office my blood pressure is always very high, as in they always freak out

All you should need to do is take your own readings in. I normally take in one reading a day for the past 7, plus my resting heartrate, plus my weight. Saves time and lets them address other needs.

The juxtaposition of clippy and AI is just too perfect.

Suppose we have an AI whose only goal is to make as many paper clips as possible. The AI will realize quickly that it would be much better if there were no humans because humans might decide to switch it off. Because if humans do so, there would be fewer paper clips. Also, human bodies contain a lot of atoms that could be made into paper clips. The future that the AI would be trying to gear towards would be one in which there were a lot of paper clips but no humans.

I wonder if the person driving it had a PhD to steer and guide and fix up code hallucinations as it "burned through tokens" for five hours to achieve this college level programming task.

I also wonder just how many tokens were burned? That is a cost thing after all.

And if not actively vibe steered by an experienced coder supervising as they burnt tokens for 5 hours at an enhanced rate, I would be impressed only if I saw the complete preprompting, and probably how many KwH of power/Litres of water it also churned through with that enhanced token burning for 5 hours.

According to ChatGPT, Ollama does make use of a random number generator for some reason.

For a very good reason. Imagine a system without any entropy from temperature. It would be quite boring. Have a standardised environment of hardware and software stack and model, set your temperature to 0 and your seed to a known number, and you will find the output deterministic and reproducable. Change your video card, pytorch version, or whatever, and you might see some drift, but as long as the temperature and seed and context remain fixed, you know exactly what's coming out.

Those with potatos still need to play, but there comes a day the game engines really need to retire compatability with the GTX750 and even the GTX1650.

I've blocked their dysfunction and chaos for years at the router. I personally found doing just a single browser isn't enough to cover all my devices and their tracking logos all over the place. Well, at least the obvious ones.

As for youtube; it's now swamped with British Robot Guy#2 and midjourney clickbate titles. Nigh on useless. Rare signed out use only for me now without cookies.

From my peer group, people are absolutely switching off, so I am starting to think the cre-ai-tors have to vibe-code their own ai-dience to watch their vid-ai-o slop for ai-dvertising clicks. All very Ouroboros. Or at least a centipede with chains of ai consuming each others crap.

...DeepSheikh...

I wish I still had modpoints for you.

I wonder if Adobe should buy Microsoft, and make Windows a monthly subscription deal!?

The true use of AI starts to show. Given the ease of searching over the old keyword engines like Google, Bing... le Chat j'pete will now use personalised language to gently steer the enthralled public directly towards a predefined merchant. The ultimate salesman.

My first job used NCR's CRAM, which were long magnetic cards with 8 hooks on one end. Each card was record-sized.

CRAM Cards were threaded onto a series of snail-cam shaped rods. When a record needed to be obtained the 8 rods holding all the cards would turn to match the binary address of the card, and release the exact card required, which fell into a hopper and was read, then put back in the random stack.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fs3data.computerhistory...

Two AIs walk into a bar. First one says "I applied for 500 jobs today". The second one replies "That's nothing -- I created 500 jobs today" They both toast to their fictional employment histories as the human recruiter in the corner weeps, sorting through thousands of perfect resumes for positions that were never real.

They should start quickly, but there will be a substantial ramp up in the number of CVEs that are caused by LLM code generation and lazy, untrained developers not looking at what is output. There will be vast style inconsistency, there will be logic failure bugs, and even the >0 chance of potentially adversarial trojan horse code. Logically, when it comes time to debug, the first human who really look s at the code will have problems far greater than someone working on untouched COBOL code from 1980.