Forgot your password?
typodupeerror

Comment Re:Risk reward and perverse incentives (Score 1) 39

This is all true and part of the reason the successful nucelar programs currently, the ones building new plants and plans for expansion and have kept up those capabilities are all State-Owned-Enterprises. The risk-reward and incentive structures just are not there for private industry.

I have always felt that if we are going to do nuclear power it should be done as a government operation. Not as a for profit enterprise, but as a public good. We successfully operate nuclear reactors for military purposes. Treat it the same for civilian needs. Military security. Strict operation and maintenance protocols. Force Majeure decisions on where and how it is operated and waste handling/disposal decisions.

Comment Re:Risk reward and perverse incentives (Score 1) 39

Bad people are gonna do bad things. That doesn't mean we should not try to live our lives anyway. It is not worth giving up our freedoms because we are afraid for our safety.

That said... I would prefer that they built new modern design reactors than restarting older reactors -because safety :)

My biggest concern with these deals is that we-the-people do not foot the bill. No tax breaks. No rate increases for citizens. Let the private industries pay for it out of their massive revenues. Google wants the energy, Google has the funds to pay for it. Let 'em. And make them pay for the regulation and security and inspections and disposal of the waste and every cost associated with running the plant. Make them set up an insurance bond to cover future costs in case something goes wrong or the operating company goes bankrupt. It is expensive and it should be. It is a big ask, but not unreasonable -just part of the cost of doing business.

Comment Re: Who cares? (Score 1) 236

You asked "what to do if your mental state depends on sunlight". I suggested going outside during lunch, when the sun should be at its strongest... that should help with general sunlight depravation issues.

If you need sunlight in the early morning specifically vs needing sunlight... then there are products you can purchase to simulate solar exposure. Lamps and lighted caps/visors that shine full spectrum light directly on your face while you are about your morning routine. They are made for people who suffer from seasonal affected disorder, and for people who work night shifts. Buy one.

Comment Re:Depends (Score 1) 236

Florida might go with permanent DST, but it's horrible anywhere else. Sunrise would be after 9am in Atlanta, and that's not exactly Northern.

OK. I have to call you out on this one. That is a trope.

With permanent Daylight Saving Time, the latest sunrise in Atlanta would be 8:45am vs 7:45am without.

If you are trying to get to work/school by 8am, you are up in the pre-dawn hours either way.

Comment Re:Fundamental physics has never been more dead (Score 4, Funny) 28

"Okay, nuclear power. Yeah. Yeah, they did give us that. That's true. "

"And industrial-scale superconducting magnets!"

"Don't forget GPS satellite relativistic corrections!"

"And radioisotope therapy!"

"Electron beam therapy?"

"World Wide Web, wasn't that created by particle physicists?"

"Yeah, they did do that."

"Capacitative touchscreens too, you know."

"All right, all right, FAIR ENOUGH! But APART from computers, telecommunications, nuclear power, PET scans, MRIs, satellite communications, superconductors, radioisotope therapy, electron beam therapy, heat-shrink tubing, cargo and baggage screening, smoke detectors, muon tomography, capacitative touchscreens, and the World Wide Web, WHAT HAS PARTICLE PHYSICS EVER DONE FOR US?!"

"Synchrotron light sources for drug discovery?"

"Oh, drug discovery?! GROW UP!"

Comment Re:Very funny (Score 1) 72

if this were true it would not be possible for the system to read the watermark.

You confuse "invisible to humans" with "invisible to statistical analysis". Humans can't, for example, see the difference between metameric colours. Statistical analysis can. Humans can't see minor drift at scales where Brownian motion dominates. Statistical analysis can. Humans can't see the difference between background disease rates and new rising threats. Statistical analysis can. On and on. Statistical analysis is a far more powerful tool than human perceptual abilities.

this is also a lie. Identical thoughts must be expressed identically, otherwise they are, in fact, *different* thoughts.

I literally gave an example above. If you saw "picked a mango", "selected and picked a mango", "from within the tree, picked a mango", you would never think there was anything going on in the choice, and it isn't going to change the story one iota.

Comment Re:Meanwhile, back at the benchmarks (Score 3, Interesting) 25

It actually makes RAM needs worse than a dense model, for a given quality.

That said, inference techniques like FreeToken are improving MoE swapping performance, so it's not as bad as it once was. Honestly, I would not be at all surprised if we start doing training in a swap-aware manner, where expert cache misses count as loss and models can queue experts to start loading before they're needed. Could even train in two tiers of loading - RAM and disk. Wouldn't that be great? Could run multi-terabyte models on consumer hardware. Training would tend to tend to concentrate key logic and reasoning on small number of very active experts, mixed reasoning/knowledge on less common experts that are usually kept in RAM, and rarer knowledge on experts that usually remain on-disk until needed.

Comment Re: Meanwhile, back at the benchmarks (Score 2) 25

There are well known, reputable abliterators on HuggingFace.

Also, censorship usually (not always) is something you care about for chats, not agentic work.

A .safetensors or .gguf model cannot run code unless you let it. No agentic framework = all it can do is write text. It can't hurt you directly. At worst, it can only lie to you.

Slashdot Top Deals

Round Numbers are always false. -- Samuel Johnson

Working...