Forgot your password?
typodupeerror

Comment Re: At least half true. (Score 1) 87

Putting out unvetted AI coded patches is no more unsafe than putting out human coded patches that aren't checked out.

Separately this is untrue given at least humans have metacognition.

Did I trigger you with the FTFY comment? Relax a bit, take a break.

As a person who has been around coding since the 1970's, starting on good old IBM360's, I've seen plenty of horrid human code that is just fail. Between bitched up code that would cause a core dump and use up huge amounts of printer paper on the 360's to Windows updates that cause endless boot cycles - It was improperly checked updates and patches.

I've also seen some Claude based code that is tremendous, much superior to the code it replaced.

How does this happen? It is vetted out, just as you would vet out any code, human or AI.

Yeah - metacognition, that's right. Same thing as checking out the patch

But perhaps more importantly, simply bitching about AI coding is not going to change it. I've seen the slop, but it won't remain slop forever. As I noted in another post here, I'm using a Software Defined Radio application written by Claude, and vetted out by humans before release. And the program is already superior to the human-written version.

Dunno what to tell ya, other than make certain you are a top tier coder, because the AI is moving up the chain of competent writing. Because you can tell me I'm wrong all day, substitute metacognition as some counterargument to "checking out", "vetting out." Frankly, I find the terms essentially interchangeable.

Comment Re: At least half true. (Score 1) 87

How are people are trusting LLM to write complex apps

I have a number of programs that were coded with Claude for the Flex radio. This is a Software Defined Radio that is a server with an RF front end.

Original software was Windows only. Decent stuff, but a lot of people really wanted Mac reliability, and Flex stood firm, Windows only. Want it for a different platform? Use the APIs and do it yourself. Then SmartSDR came out for MacOS. A better program, so okay.

A group using Claude took the Flex API's and AI'd it. It was done incrementally, and all implementation and changes were reviewed and tested by actual humans. No rewriting everything from scratch every addition.

The results? More software, free software, and some of which are demonstrably better than the human only coded versions. One in particular, AetherSDR, is now a daily driver. MacOS, Windows, and Linux. And pretty lightweight - I've run it on a RPI 5.

AI coding is here to stay, and with proper implementation and review, can work well.

Comment Re: At least half true. (Score 2) 87

(trusting) AI generated patches to AI generated problems (both real and hallucinated) is a recipe for disaster.

FTFY.

Putting out unvetted AI coded patches is no more unsafe than putting out human coded patches that aren't checked out.

I think some groups trust AI just like they trusted early GPS guidance. People lost their lives when early guidance had them drive off bridges or lost on trails in the desert.

Comment Light and radio waves [Re:Power beaming] (Score 1) 109

Light and radio waves are both electromagnetic radiation, but the difference is frequency. Light has a frequency something like 100,000 times higher than radio waves (depending on the frequency of the radio waves, of course.)
So there really isn't any way to make a laser at radio frequency-- if it's radio frequency, it isn't light, and lasers are Light Amplification by Stimulated Emission of Radiation.

The difference is just a matter of definitions, of course, but we define the two using different words because the difference in frequency of factors of 10^5 or more really make the behavior of light very different from the behavior of radio waves.

Comment Re:The 0.01% of exceptions Waymo has not hit yet (Score 1) 53

"I don't need to outrun the bear; I just need to outrun you!"

Human drivers are so incredibly terrible and deal with those driving exceptions so badly that by comparison, Waymo ain't so bad.

So why are humans still allowed on freeways? Seems like making perfect the enemy of good. Waymos are safer, so they shouldn't be the only ones left on the freeways. That is logic 101. Waymo safe, humans unsafe. Simple.

Comment Re: Good (Score 1) 92

I'm waiting for the keeping-urine-in-jars stage of his Howard Huges-ness, myself.

One of the hidden dangers of mental illness is that a person can be insulated from their madness. If we step away from Musk for a moment, we can look at Michael Jackson. it was pretty obvious to most people that he was a pedophile, and liked little boys. He also was so addicted to facial surgery that he disfigured himself. He took drugs that eventually killed him. But his money insulated him from his madness, and Jebbuz K Ryste, people supported him and even brought their young boys to him.

Howard Hughes had severe OCD and depressive symptoms -some have called him schizophrenic, but the jury is out on that. His wealth almost certainly made things worse. Now Elon's particular aberration is based on drugs like Ketamine. He supposedly takes small doses daily, and larger doses to party. There was a huge change in his persona a few years back. He was once middle of the road, and a bit to the left, and suddenly he went full MAGA. That generally does not happen, people might adjust their views, but his change was drastic.

But he has a cult who worships him. And while he feeds off that cult, it eventually harms him His money and his cult protects him at the moment, but can only do that for so long.

I'm sure I'll come under attack for the blasphemy, the cult is very easy to trigger - one guy who is playing the goalpost moving game because I dared to note that Spacex lands some of their rockets, but treats satellites as disposable. That has degenerated into claims of my incompetence regarding orbital mechanics. Well, okay, I have some random on the internet all spun up.

Comment Re: Good (Score 1) 92

I didn't say there was, and you still seem to be ignorant about it. There's no rhyme or reason for external de-orbiting of Starlink birds.

Well now, I made a simple statement, that retrieval of defunct satellites is a good thing, it appears that you took offense to it. To the point of claiming I know nothing about orbital mechanics, and apparently everything I have written is simply wrong. Or off topic.

Space junk is never off topic in discussions about space.

Of course not every example I gave was directly focused on Starlink Sats. I was discussing the issues involved with Sats, we can expand that to other space junk.

There is a reason for that. Not every defunct satellite in orbit around earth is a Starlink satellite. But most are. Around 11,000 in October 2026.

Even if we don't return them to earth, getting them out of orbit is a big plus - there might be 170 million pieces of junk orbiting, defunct sats, rocket stages and other things that can make for a bad day. There are many groups working this problem. Including the European Space Agency. https://www.scienceabc.com/pur...

https://www.esa.int/missions/a...

https://www.astroscale.com/en/...

So relax a little before returning to the insults. Fact is, there is no physics to keep us from capture and return. There is no physics to keep us from capture and rapidly deorbiting. There are groups actively pursuing this, and clearing out dangers debris, as I have cited.

One thing that many do not understand is that collisions spew out debris well beyond the present orbit. While not a collision event, and as an example, on February 2015, a Defense Meteorological Satellite Program (DMSP) satellite of United States experienced a single breakup event induced by battery explosion, the resulting debris were so widely dispersed that extended more than 300 km above and below the pre-breakup orbit. There is nothing to prevent that happening in a collision either. Pieces go flying off in different directions and velocities. The effects are essentially unpredictable, but real.

https://www.sciencedirect.com/... "Satellite breakup behaviors and model under the hypervelocity impact and explosion: A review"

You'll need academic access, but you probably do already - so you can read it for yourself. If you wish to correct the authors, there are mechanisms that can lead from simple correction to forcing a retraction.

This is important stuff, critical to military applications and defense. Spacex as well as everyone else. It affects launch envelopes and mission profiles, and has mission ending consequences. Collisions have much wider effects than just one orbital shell. That's why people are working on it. It is important.

And if you consider this another example of my lack of chips in orbital dynamics, you can simply ignore me.

Comment Re:More than 3 possible choices [Re:Bad... from a. (Score 1) 94

And in fact we are seeing a mix of options, including solar plus storage. Spring 2026 Solar and Storage Industry Update

That link appears broken, got something else?

I clicked it and it worked for me. There were various internet outages yesterday, you may have hit one.

Since the cost of power is one important consideration for the location of data centers, they have some latitude to bias their site selection toward places with cheaper power.

When it comes to selecting a site for data centers and solar power I see a contradiction that is not likely to be easily resolved. Solar likes dry places, wide open remote places to keep land prices low, among other constraints that may or may not be relevant.

It's one factor in siting. Not the only one. Sites do exist.

As I see it there's nobody "leaving out" solar.

Other than you. You are looking at the solar through the lens of the technology of ten years or more ago.

...
I expect solar to stay in that place for a long time.

You need to update your crystal ball.

...
Solar PV and batteries? We've been trying to do that since around about 1980, when Jimmy Carter put solar panels on the White House and told everyone to wear sweaters. How's that worked out so far?

And that is the perfect example of the problem here: you are citing technology from 47 years ago and thinking it has something to do with today.

Your knowledge is simply out of date.
  https://images.fastcompany.com...

Comment Re:NORM!! (Re:Welp guess they'll do AI in China) (Score 1) 94

Perhaps you can enlighten me, and the rest of the class, on how things really work. Calling thorium rich tails something like weapon grade plutonium may be a bit hyperbolic

Correct. And that's where I stopped reading. When you start out by showing your rhetoric is not constrained by actual reality, I don't see the point in reading further.

Comment Re:More than 3 possible choices [Re:Bad... from a. (Score 1) 94

Except that there are more than three options. You leave out, for example, solar energy with storage, which is now cost-competitive in many areas of the country. (And other renewable sources, of course).

If that were true then we'd not be seeing plans for nearly 200 GW of new natural gas fueled behind-the-meter electricity production. Rather we'd see 200 GW of solar+storage, or more likely a mix of options

And in fact we are seeing a mix of options, including solar plus storage.
  Spring 2026 Solar and Storage Industry Update

because as you say solar+storage is cost competitive in "many" areas than "most" or "all" areas.

The fact that the cost of power from solar generation is site dependent seems to be missed by far too many slashdot commenters; both the pro-solar and the anti-solar factions. The pro-solar ones say "look! solar is cost competitive [pointing to data in one spot}! Therefore, solar is the solution!" The anti solar ones say "look! solar isn't cost competitive everywhere! Therefore, solar is not the solution!" Both are wrong.

Since the cost of power is one important consideration for the location of data centers, they have some latitude to bias their site selection toward places with cheaper power.

It would seem that you, and others that think like you, are failing to comprehend the scale of what was in the fine article. The planned 180 GW of new natural gas electricity production is on the level of all existing solar power production on the grid.

In 2025, between 607 GWdc and 698 GWdc of PV were added globally [reference]

It took us decades to get that much solar on the grid, just how quickly do you believe we can double that so data centers can get the behind-the-meter solar+storage production they believe they need? Where will all these new PV cells, batteries, and so on come from?

Turns out the lead time for solar, including solar plus battery, installations is comparable to gas turbine power installations, and of course vastly superior to nuclear.

I leave out solar+storage because that is still an unproven technology.

About all I can say is that you are behind the times.

You say using less energy means higher prices

No, I say PRODUCING less energy means higher prices.

OK. In that case, I will only repeat my original statement: analyses conclude that improving energy efficiency (i.e., using less energy) is in fact far more cost effective than building more power plants.

Comment Re:So short term we took a hit to grades (Score 3, Informative) 21

Unfortunately funding has never corresponded to improved performance in schools beyond some minimal threshold. Kansas City learned this decades ago and NYC charter schools that have better results for less money on average. A monopoly on anything never produces the best possible results and I'll never understand how the same people who will rail against private industry potentially having one will believe that an actual government one will magically produce desired results.

Your entire post is utterly stupid (and factually wrong) because the Federal government has very minimal control over primary education which is largely controlled at the state and local level. The Department of Education didn't exist until 1980 and has done nothing to improve educational results during its existence. Taking it out behind the barn and putting it down fundamentally changes nothing other than making a bunch of charlatans find another grift. Consider a city like Baltimore that has abysmal school performance despite being in both a city and a state which have voted solid Democrat for decades and explain to me why this is somehow the fault of evil Republicans in the federal government despite the funding and policy being almost entirely at more local levels.

Comment Re:NORM!! (Re:Welp guess they'll do AI in China) (Score 1) 94

If there was "too much" (in scare quotes since this is entirely arbitrary) NORM then the mining tails were considered something hazardous on the level of weapon grade plutonium.

I stopped reading here. You clearly don't have even a hint of a clue about the restrictions on weapons-grade plutonium, and given that you show your ignorance in the second sentence of your post, no point in reading further.

No, mine tailings are not considered equivalent to weapons-grade plutonium.

Comment Re:Michigan doesn’t run the economy. These d (Score 2) 94

What gets lost in these stories is that data centers don’t burn electricity for its own sake. They run the websites, databases, payment systems, logistics, and yes, AI that nearly every US business depends on. This isn’t just Instagram and TikTok; it’s payroll, inventory, banking, and healthcare records.

Nope.

All that was already being done by existing data centers. The new hyperscale facilities are being built for one purpose: AI.

Not running small businesses and serving TikTok videos.

Slashdot Top Deals

Always look over your shoulder because everyone is watching and plotting against you.

Working...