We have used water before lithium, and it isn’t bad at all.

Not so great in a flat dry desert though. Pump storage is great when there is lots of water and a naturally occurring elevation, but there’s lots of places on Earth that don’t have that, but do have energy to store.

This is big! Grid scale Sodium Ion battery technology is (on paper) the best candidate for cheap large scale electricity storage. The fact that this company is working on 9 pilot deployments mean that this will likely produce the real world results that the paper exercises promise.

There are SO MANY advantages of Sodium Ion battery tech for grid storage over everything else we’ve used so far (nearly all Lithium based).

Sodium Ion batteries:

• don’t have as intense thermal management needs Lithium chemistries
• don’t have the massive negative environmental impact for their source materials (because its a part of regular old table/sea salt)
• doesn’t have the massive swings in capacity when operated in extreme hot or cold temperatures. Sodium Ion doesn’t care.

The only downsides to Sodium Ion is that the batteries are physically larger for the same amount of energy stored (which isn’t a problem for stationary storage), and the charging/discharging curves are not as linear as other chemistries (which again, isn’t an issue because these are purpose built applications where the curves can easily be managed by battery management systems).

Oh, and Moonlighting!

Did the car try to stop and fail to do so in time due to the speeding, or did the car not try despite expected collision detection behavior?

From the article, it looks like the car didn’t even try to stop because the driver was overridden by the acceleration because the driver had their foot pressed on the pedal (which isn’t normal during autopilot use).

I feel like calling it AutoPilot is already risking liability,

From an aviation point of view, Autopilot is pretty accurate to the original aviation reference. The original aviation autopilot released in 1912 for aircraft would simply hold an aircraft at specified heading and altitude without human input where it would operate the aircraft’s control surfaces to keep it on its directed path. However, if you were at an altitude that would let you fly into a mountain, autopilot would do exactly that. So the current Tesla Autopilot is pretty close to that level of functionality with the added feature of maintaining a set speed too. Note, modern aviation autopilot is much more functional in that it can even land and takeoff airplanes for specific models

Full Self Driving is just audacious. There’s a reason other companies with similar technology have gone with things like driving assistance. This has probably had lawyers at Tesla sweating bullets for years.

I agree. I think Musk always intended FSD to live up to the name, and perhaps named it that aspirationally, which is all well and good, but most consumers don’t share that mindset and if you call it that right now, they assume it has that functionality when they buy it today which it doesn’t. I agree with you that it was a legal liability waiting to happen.

Don’t take my post as a defense of Tesla even if there is blame on both sides here. However, I lay the huge majority of it on Tesla marketing.

I had to find two other articles to figure out if the system being used here was Tesla’s free included AutoPilot, or the more advanced paid (one time fee/subscription) version called Full Self Drive (FSD). The answer for this case was: Autopilot.

There are many important distinctions between the two systems. However Tesla frequently conflates the two together when speaking about autonomous technology for their cars, so I blame Tesla. What was required here to avoid these deaths actually has very little to do with autonomous technology as most know it, and instead talking about Collision Avoidance Systems. Only in 2024 was the first talk about requiring Collision Avoidance Systems in new vehicles in the USA. source The cars that include it now (Tesla and some other models from other brands) do so on their own without a legal mandate.

Tesla claims that the Collision Avoidance Systems would have been overridden anyway because the driver was holding on the accelerator (which is not normal under Autopilot or FSD conditions). Even if that’s true, Tesla has positioned its cars as being highly autonomous, and often times doesn’t call out that that skilled autonomy only comes in the Full Self Drive paid upgrade or subscription.

So I DO blame Tesla, even if the driver contributed to the accident.

I mean, we’re all posting and conversing about this on Lemmy. Isn’t that exactly what happened already?

I see a couple missing from your list off the top of my head from the 80s:

• Hardcastle and McCormick
• Simon and Simon

The workarounds are interesting: link

Both workarounds would appear to do stuff that retriggers HAL detection. So something about the config must get broken, and only HAL redetection is needed to bring the VMs back to life. Now I’m really curious what the underlying bug/problem is.

Ah yes, the Lego caltrop.

First, it looks like this may be a dressed up advertisement for their newly released book:

My book on Enshittification is coming out in a couple of months, and the early reviews are already coming in, and they are gratifyingly glowing.

That fact alone doesn’t discount their argument, but it should be considered.

Second, I disagree with this premise of the author:

Because this isn’t an individual problem, it’s a systemic one.

I disagree, its both.

As the author rightly identifies, there are somethings that are only addressable systemically such as healthcare of mass transport. However a whole other host of items the author references are absolutely individual problems. Example from the author:

When all your friends are going to a festival, are you really going to opt out because the event requires you to use the Ticketmaster app (because Ticketmaster has a monopoly over event ticketing)?

Yes, I opt-out of nearly every Ticketmaster event. It is an individual problem with an individual solution.

If so, you’re not gonna have a lot of friends, which is a pretty shitty way to live.

My friends largely also opt out. Perhaps we self select for like-mindedness.

This means that they don’t have to worry about losing your business or labor to a competitor, because they don’t compete.

They can still lose my business if I opt out of the entire industry, such as corporate social media. No amount of competitors changes my mind on that. This could also be done on streaming services, choosing to read instead etc.

This isn’t just a systemic problem as the author suggests.

Kinda surprised me, actually. Red Bull has been pretty poor performer recently in spite of Verstappen’s amazing talent. RB doesn’t have Newey anymore for aero, and won’t have the Honda power plant for 2026, but all the rule changes for 2026 mean RB can’t ride on existing completed legacy work that folks aren’t there anymore to recreate in 2026 rule conformity.

If Max wanted to make a jump he’d have a whole bunch of good reasons to do so. Toto has as good power plant, a very competitive car, a huge chequebook, and a still stinging lesson to listen to their star driver when changes are needed.

How many people have said “Just imagine how good Max would be doing right now if he were in a good car!”

He would have done well with the silver arrows, I think.

I was told it was originally pitched as his character from ER but they ended up having to re write it cause they couldn’t convince whoever they needed to to use the ER production.

There was a lawsuit about this from the widow of Michael Crichton (who would have had to sign off on ER IP). The widow claimed The Pit is an ER ripoff and that the Crichton estate should be compensated.

Are there any AI services that don’t work on stolen data?

Yes, absolutely, but I don’t think that’s the question you want the answer to. There are many places where AI is used inside companies or hobby project where the specific problem to be solved is very specific and other peoples stolen data wouldn’t help you anyway.

Lets say you’re a company that sells items at retail online, like a Walmart or Amazon. You want an AI model to be able to help your workers better select the size of box to pack the various items in for shipment to customers. You would input your past data for shipments you’ve sent including all the dimensions of your products you’re selling (so that data isn’t stolen), and input all of the sizes of boxes you have (they’re your boxes so also not stolen). You’d then could create an Unsupervised Classifier AI model based on linear regression. So the next time you have a set of items that need to be shipped out you’d input those items, and the model would tell you the best box size to use. No stolen data in any of this.

Now, the question I think you’re asking is actually:

“Are there any LLM AI chatbot services that don’t work on stolen data?”

That answer, I don’t know. Most of the chatbot models we’re given to set up chatbots are pretrained by the vendor and you simply input your additional data to make knowledgeable on specific niche subjects.

Along those same lines we also have the classic: how is prangent formed

We already have examples of this, with some deep meta.

The TV show MASH depicted an army field hospital from the Korean War in the 1950s. They frequently called out specific medical procedures for treating various conditions. One of the main actors was Alan Alda who played a lead surgeon. Many years later, the same actor Alan Alda again played a doctor on the medical drama ER. I remember there was a scene where they needed to treat a specific emergency medical condition and there wasn’t the time/materials/whatever to do the modern technique one would use for this. Alda’s character suggesting the really old method (that was used as a contemporary technique his character used in MASH). I can’t remember if they ended up using it or if they shot down Alda’s ER character for the method being out-of-date.

Cool, if the storm is over, then they can release the unredacted Epstein files now, right? They should have no problem doing that.

Make it a reverse origin story. The movie opens with the reanimated horse and the necromancer girl as a funny “fish out of water” story leaving us wondering WTF is going on to arrive here, and only in the third act do we get the reveal of what actually happened with our first view of necromancer girl as an absolutely obsesses horse girl.

My expectation was a crude/debauched way of interacting with books. “I’m gonna dogear the pages.”

Right, that’s way too literal (pun not intended). Too superficial. Too easy.

A more nuanced interpretation of the joke than is apparent requires so much relatively heavy-lifting from the audience and/or for the audience to work backward from the decision that “the joke must be competent/fulfilling”.

Yep, there are several leaps of logic the audience must go through to arrive at the author’s position. Hence it being much more satisfying when we get the joke. Moreover, there’s also a transference of appreciation, and perhaps a bit of envy on the part of the audience that the “her” in this case has such a loving mate that we would all want the same.

I have zero faith this campaign will stop until we’re back in the stone age.

Future headline:

“Game vendors pressured to stop selling games where women are allowed to vote or express an opinion that doesn’t reflect her husband’s”