Sam Altman has claimed a lot of outlandish things, but he's never said AI would solve all the world's problems, take us to the stars, stop world hunger, fix our governments, or save the planet. YOU made all that stuff up.
"We are close to creating the genie that can grant any wish." ~Sam Altman
If you haven't used AI in a meaningful way in the past, oh, 6 months, you should rightfully consider your opinion invalid on this topic. I know it's popular, particular with "real programmers" to make these hubristic statements, but there's no truth in it. It's usually those most heavily steeped in the old ways of doing things who have the hardest time with new things.
The same was true 6 months ago. The field has changed a lot, with breakthrough changes almost daily - so much so that just keeping track of everything released is an exercise in and of itself, nevermind testing it. I have hardware set aside just to do that, and there's often a 2-3 day queue.
Having used AI for years the best compliment I can give is at least the latest GLM-5 models are useful enough to find bugs. It often takes a day to process and the vast majority of the output is either wrong or incoherent gibberish but at least it still works. This talk of daily breakthroughs has not been my experience. Have franken models from 2023 I still use to this day. This wouldn't be the case if it were all breakthroughs every day.
Throughout that time it has been a similar level of limitations and nonsense. The primary difference I see today stems from focus on post training CoT behaviors after the pretraining well had mostly run dry.
Would you consider a facility producing clean water and power (like a nuclear desalination facility) to be "snake oil"? You know: things which are in high demand and improve the lives of many. No, AI isn't clean water or power (and it does use those things). But it is akin to a revolutionary change, like the steam engine, or the Sterling engine, which completely changed the face of industry in a very short period of time.
Personally, I do hold it against people if they're not significantly using AI at this point. It's a self-enforced personal limitation in capabilities.
Personally, I just wish AI users didn't themselves sound like AI (or Sam Altman).
You still have to work, but it's hard to argue against the economics of it: a single developer can 10x output with ease,
I don't understand the 10x output assertions. If anything even remotely like it were true it would be noticed in worker productivity data. This simply isn't happening. One would be hard pressed to find any signal that diverged substantively from the framing trendline.
and the quality is significantly better than most developers can produce.
So now AI is better than most developers?
It speeds architectural design significantly and improves associated outcomes, even when you know what you're doing. It helps identify faults.
Maybe if you run a code mill or do a lot of rote boilerplate work AI could automate churning out code. More likely all this is doing in the end is creating suboptimal paths of least resistance you will pay for later in the product lifecycle.
I've yet to face a situation where rate of "output" itself was a significant factor in the overall effort. While sometimes I wish my fingers could move faster yet the underlying constraint is human thought to design/organize systems.
Even outside the domain of programming it's a game changer in many ways,
While LLMs are indeed useful it doesn't excuse the peddling of hyperbolic nonsense.
though we're just now only starting to get to the point where those capabilities can be realized.
This has been a constant fixture of the last few years... this constant drip of amazing new model with amazing new capabilities time and time and time and time and time again... so amazing it changes everything. So amazing it has to be regulated and kept from public view for fear of triggering human extinction. We have all listened to this same tired crap repeatedly for several years now. All of these amazing just now unlocked capabilities that just today are changing everything for real this time.
I do wonder what the Venn Diagram is of people who are running CPUs and systems which are older than 18 years but equally need to run the absolute latest OS version.
I hope normal distros don't start pulling this shit. Have an Atom server circa 2009 that won't even boot if this sort of crap started happening. It's not like you can elect not to update because they stop updating older versions.
Also have a laptop of the same vintage I use regularly... perfectly good hardware even by today's standards. Completely unnecessary, wasteful and pointless in the current environment with absurd hardware pricing. There is no reason they can't just use dispatching to exploit new features as has always been the case.
i said safer, not bullet proof! Even a lead/acid battery is dangerous if you do stupid things!!
What you said is flat wrong. Sodium is NOT safer than LFP used in nearly all present day power stations and residential ESS. It is not the same as cell phone batteries as you asserted.
The companies selling these products like to imply that you will be able to power your whole house with their systems. Guess what, they can't. Because most US panels use split phase to break the 240V incoming into 120V circuits, you only get your plug-in solar feeding half the breakers in your panel. Now, in order to do plug-in solar properly, you should have a dedicated plug for the inverter. While you have the electrician there, have them arrange the breakers on your panel so your large draws (AC, washing machine, etc) are on the same leg as your dedicated solar plug.
A solar system does not produce 240V.
It produces 24V or 48V. Of course you could wire it to make 96V
However: DC has some issues
Solar panels output DC and are typically run in series strings of hundreds of volts. Most PV panels are rated for series voltages of at least 1000V although residential systems tend to max out at under half that. For balcony in the US I think everyone will keep voltages under 80V per MPPT to align with NEC 690.
The 120V circuits parent was referring to are AC output of power electronics that convert DC from panels for grid/residential service.
I'm surprised.
We're living in a crazy world.
Yes they have.
Just not in Gods own Country.
Please stop making shit up.
Sodium isn't the one weird trick that leads to a step change in energy storage.
It is. Battery prices dropped by nearly 90% recent weeks.
No they haven't.
Except it very much is. Within any locale the government dictates names.
The GNIS has changed the name so any map based in the USA not using the updated Lake America is objectively wrong.
The government can decide all it wants what it wants to call something. Nobody else is obligated to care. Calling a label on a map the government doesn't control "objectively" wrong is based on a fundamentally incorrect assumption of what a map is.
I agree. But we don't live in a world of "shoulds". We live in a world of laws.
We live in a world of legitimacy. Laws are merely an abstraction bolted on top. Not that your statement is even relevant in the first place since map makers have no legal obligations whatsoever to parrot government labels.
The president should not have the authority. But he objectively does have it. Also he can't change the name outside of America, the ruling only applies to what America calls these places.
Again the president has no authority whatsoever to tell map makers what to do.
The only reason to adopt someone else's naming convention is politeness.
How does politeness apply to changing the name of Lake Ontario? Was it done to be polite or was it done out of pure unadulterated spite?
Do you still call it Turkey in English or do you use their preferred name Türkiye? I'm guessing you don't even know how to write an umlaut on your keyboard.
Personally I'm willing to compromise on Taqiyya.
I'm taking the side that the naming system has an authoritative source. Or are you in favour of your ISP sending you to different addresses for domain names just because they should feel like it?
The premise is incorrect. Maps are not like DNS.
And for the idiot terrified of trans-women in women's sports, how does he - I'll say their pronoun is 'he' with 99% probability - feel about trans-MEN in men's sports?
The only reason segregated sports exist is men and women are physically different. Calling a man a women and allowing them to play in woman's sports is cheating. In other words it is not fair to actual women who are at an inherent disadvantage to physical males. If someone wants to go in the other direction and accept a disadvantage over the other players that is a different matter entirely.
Again the whole point of segregated sports is **discrimination** by physical characteristics not self image or pronoun preferences.
The advantage of sodium battery means anybody can produce it, there's no geopolitical alliances needed to manufacture them. Lithium is super rare by comparison.
Lithium is not a substantive impediment to battery production. Lithium is cheap and abundant enough supply is never going to be an issue short of lead time required to respond to increased demand.
Production of sodium batteries requires substantially similar industrial processes to LFP. Anyone with facilities to crank out batteries at scale is sure as heck not being controlled by sourcing of lithium. Other required inputs such as graphite for hard carbon production are far more concentrated in a market dominated by none other than China.
Any and every country can domestically produce sodium batteries forever.
This notion all you have to do is replace lithium with sodium and then everything changes is not the case. Whether you are producing LFP or sodium doing so at scale at competitive cost and quality requires similar capital investments as for LFP.
The big benefit is any country can stick a straw in the ocean and build a battery factory regardless of sanctions. Go look at Iran right now if you want to understand why I keep harping on this point.
Again whether you pick lithium or sodium neither is a significant cost center or impediment in the overall production effort. Sodium isn't the one weird trick that leads to a step change in energy storage.
Not just about cost, Lithium have better capacity both by size and kg, but for home energy storage, that isn't a important factor. Lithium batteries are more dangerous and with limited charge cycles... and this is important for a home energy storage... a cell phone burning is dangerous, imagine a several KW storage system!!
All of the current residential ESS use LFP which is much safer and has different properties than cell phone batteries. They can be punctured, over and undercharged and operate outside of operating range without risk of the thermal runaway of the li poly cells in a cell phone. All of these high current electronics, storage systems and even solar panels still present a real fire risk.
As for this notion of sodium being safer... there was a video of someone shooting a Bluetti sodium power station when it was off and not plugged into anything. It at first slowly began to smoke, deformed, caught fire and went up like a flame thrower + fireworks show.
I dunno, people on boats and in vans buy this stuff and run it for a decade or more without problems.
What stuff? What are they buying? Power stations or discrete components for a low voltage system? The technology in the form of modern power stations didn't even exist "for decades"... Ecoflow the company is less than a decade old. Decades ago you bought a big honkin LF inverter and connected it to lead acid batteries. Completely different technology.
Name brands like anker and ecoflow have very solid track records and 5-8 year warranty on most of this stuff.
Anker and Ecoflow don't offer any warranty exceeding 5 years on their power stations.
It's not fly by night Chinese crap. All this stuff has a very solid track record at this point. I bought my first MPPT way back in 2017 it's still alive and kicking no problem
You purchased a discrete MPPT not a power station. Most of these companies won't even sell you a replacement board if the MPPT is blown.
Hard work never killed anybody, but why take a chance? -- Charlie McCarthy
