> Also, these assistants (for now) appear to exhibit no common sense about what is “much”, “little”, “exceptional”, “average”, etc. For example, after measuring a consumption of 3.5GB of memory (!!) for solving a 3-disk problem (due to a bug), the assistant declared all was well...
These figures are very plausible. Most Linux distros are terribly inefficient by default.
Linux can actually meet or even exceed Window's power efficiently, at least at some tasks, but it takes a lot of work to get there. I'd start with powertop and TLP.
At least on Thinkpads over the years, I've never seen anything remotely close to that either. I've had my Thinkpad x260 power draw down to 2.5 watts at idle, and around 4 or 5 watts with a browser and a few terminals open. That was back in 2018! With the hot-swappable battery on the back, I could go for 24 hours of active use without concern.
A few years ago I was playing around with upscaling options in ffmpeg. For my starting point, I used my DVD of The Road Warrior that I bought in 1999, and wasn't particularly well mastered. I applied some filters to remove film grain, and raised the frame rate to 120 fps by inserting artificial interstitial frames.
Firstly, the filter that removed grain from the film also removed grain from the road, the sand, and Mel Gibson's stubble, all of which there's a lot of in the Road Warrior. Everything looked quite a bit too clean.
But the super high frame rate gave the video a hyper-realistic quality. Not realistic in the sense that I'm watching actual post-apocalyptic survivors. Realistic in the sense that I'm looking at what are clearly actors wearing costumes, and it's hard not to imagine the camera and rigging crew standing just out of frame.
An interesting exercise, but not how I want to experience that movie. Having said that, this was my experience just playing around with ffmpeg on my desktop PC. I wouldn't rule out the possibility that a dedicated professional using the right tools (presumably also ffmpeg) could manage a set of adjustments and upscaling processes that really do create a better experience than the original film.
I must be a few years older than you are. It was the original Cyberpunk (set in 2013, published in 1988) that did it for me.
One of the things I remember about the game was that it came with a suggested book and film list. Reading all those books, and tracking down the recommended films was something of a quest for me and my friends. That last part sounds trivial, but if your local video rental store didn't happen to have a copy of 1982's art-house weirdo indie film Liquid Sky, it was a real challenge.
When I bought my Prius in 2010, the longevity of the traction battery was a concern. A month ago, when I finally sold it to a garage, that car had over 190,000 miles on the odometer. I sold it because the transmission needed some repair work. The battery and the engine were still going strong.
To be fair, I’m pretty sure the 2010 Prius uses a completely different battery chemistry than modern electric cars. (I think it’s NiCad, but I’m not 100% sure on that.)
NiCads suck, you find NiCads in like old AA rechargeables and cheap toothbrushes.
I don't know for sure if NiMH last longer than Li-Ion but I've had much the same experience with my Prius - Old as hell and everything but the battery failing
My prius (2006 model) finally had the traction battery (NiMh) start to loose modules at about 250K miles. It was clearly getting weaker, but drivable at that time. Then Covid hit, and it sat for 2 months without being driven / charge cycled. That pushed it over the edge.
That isn't predictive at all of NMC or LFP chemistries though (and I'm not going through multiple charge cycles per drive), but a fun anecdote. It was an entertaining project opening up the battery pack and identifying/replacing the bad modules.
In the end, other parts of the car were dying too, and the final straw was California's refusal to allow aftermarket catalytic converter replacements, and the Toyota's price (with no competition) was more than the vehicle was worth.
So far my two EVs, both NMC chemistry (Kia and Rivian) are at 80,000 and 30,000 miles respectively, with no noticeable degradation.
My Prius is also a 2006 but with only 194,000 kms. The first battery module failed 21 months ago. I bought a used one on eBay for $35 and it's been running perfectly fine ever since. I keep it undercoated but the body is going to go before anything else because we salt our roads in the winter.
Same for ours, a 2010 model. It's 15 years old and the battery and electronics work fine, even the GPS system. It's the other stuff that needs fixing: A/C, exhaust, various pumps, etc.
You're right to point out the difficulty in getting projects accomplished in the face of intransigent environmental concerns. But you're also making a strawman argument. This isn't the possibility of "a bat being injured." This is more like the possibility of a subspecies of bat becoming eradicated by destroying their habitat.
To be clear, the benefits to high speed transit are probably worth destroying some habitats, and we need to weigh the social and economic benefits of allowing some level of environmental disruption. Progress comes at a cost. We should be clear about what that cost is.
As I understand it, the issue currently is that there's not really a framework for justifying or balancing those costs. Right now, the bat conservation people point out that the route will potentially eradicate a particular subspecies of bat. That gets sent up to the planning team, who now attempt to figure out a solution that prevents that from happening, and figures out how much that will cost. But what you probably want in between those points is someone to decide how much is too much to spend on the bats.
We do this for other stuff - for example, the NHS in the UK has a system called QALYs (Quality Adjusted Life Years) which represent how much a particular treatment will extend someone's life, adjusted by the quality of that life. You can then calculate a cost per QALY for a new treatment, and make a decision about what costs are worthwhile for the NHS to pay for, and what aren't.
Something like that that could apply to planning permission decisions would be very useful for national infrastructure projects.
For me the contradiction is simply how much of the status quo doesn't have to justify itself with those rules.
It may seem like an improvement to say that a rail project has to be careful about bats. But left unsaid is that the highway that already exists and competes with rail was never asked to perform such an analysis, and it's likely someone could adding lanes to that highway with much less stringent requirements.
So what is ostensibly environmental law, really ends up being a status quo law - if the status quo is bad for the environment, the law perpetuates it. The headline is about bats and trains, but everything from insects, to animals to people are killed right now - every day - on highways and no one bats an eye.
This argument is undermined by the malign behavior of green activists and their academic allies, who have been caught in the past inventing non-existent new sub-species specifically so they can use the endangered species argument to block construction projects.
A notorious case is the snail darter, invented to block construction of the Tellico Dam in Tennessee. It was the first legal test of the US Endangered Species Act and it was fraudulent [1]. This raises huge questions about how the Endangered Species Act is actually being used, if the very first test case was about a species that scientists now think doesn't really exist as a separate thing at all.
Another case is the California Gnatcatcher, which is not an endangered kind of bird, but green NIMBYs argued under the ESA that the coastal California Gnatcatcher was a different species that would be endangered by construction. They have successfully kept the "coastal" variant of the bird listed as an endangered species for decades, which regularly blocks or slows down construction in CA.
They love this game! After all, what defines a species? It's a vague concept and the taxonomists who decide whether something is a new species are academics, who are all on the far left. Nothing stops them publishing a paper that concludes the animals next to any planned project are unique and special snowflakes that must be protected, purely because they just want to block progress.
To put the scale of this problem in perspective, last year taxonomists "discovered" 260 new species of freshwater fish alone [2]. They claim that hundreds of unique kinds of fish escaped notice for centuries, that this happens every year, and each one of those kinds of fish is critical to preserve. Is this plausible?
> have been caught in the past inventing non-existent new sub-species
That's not what happened, and your own links fail to support your narrative. It was genuinely believed - before the passing of the act - that the recently-discovered snail darter was a distinct species. It is still disputed whether it's a distinct species, unique sub-species, distinct population segment, or none of the above. The first three of these would still afford it protections under the Endangered Species Act.
Unfortunately what people say they believe doesn't really matter. These decisions are so subjective that nobody can ever prove good or bad intent. And having nice intentions isn't worth much, what matters is outcomes.
The outcome here was the dam was blocked despite being nearly fully constructed, the decision was objected to by the developers who said it was nonsense, the dam builders were correct and the taxonomists were wrong. 100% bad outcome: everyone loses.
Unfortunately it's what you'd expect from a system that allows people who have no incentive to say yes overrule anything they don't like, because if later it's revealed they were wrong they can just say "whoopsie, well who is really to say what is true anyway, times change etc". There's no accountability.
> They claim that somehow hundreds of unique kinds of river/lake fish escaped notice for centuries, that this happens every year, and each one of those kinds of fish is critical to preserve. Is this plausible?
I don't disagree with the premise that environmental laws are being abused by people with ancillary agendas, but arguably humans have been underestimating the complexity of ecological systems for generations, incorrectly assuming that we understand everything well enough to manage and control our impacts without long tail side effects.
It may well be plausible that there are many hundreds of undiscovered fish species out there, and that interfering with any subset of them could cascade into other consequences. We've certainly been learning a lot about the impacts of dams on fisheries in recent years—changes made centuries ago that had profound long term effects on our food supply, to take a tangible example.
There was probably motivated reasoning with the snail darter, but to call it outright fraud is a bit rich. The hypothesis was proven wrong with technology we didn't have access to in 1973. (Re [1])
It's also kind of funny to me that "all academics are far left". I'd love to hear more about that.
Regarding what is critical to preserve, I don't know. Maybe some laws should be changed or amended.
There's lots of info about academic far leftism out there. This article is a decade old now (the problem is surely now far worse) but the trend in the graph is clear:
That's an average of self reported views, in a culture where everyone tells each other they're totally moderate and neutral. The problem is less serious in engineering subjects. In non-engineering subjects, far leftism is pretty much homogenous.
> To be clear, the benefits to high speed transit are probably worth destroying some habitats, and we need to weigh the social and economic benefits of allowing some level of environmental disruption. Progress comes at a cost. We should be clear about what that cost is.
Do you have a sense of how to approach the question of “cost” for this particular bat case?
When destroying an entire habitat (let’s assume we can define the boundaries of that habitat and it is mostly unique), do you have a sense of how to compute the cost given the multitude of species, geographical feedback loops, etc.?
Much more measured and thoughtful than I would have been, but I think you're exactly right. I don't know the first thing about bats but even I know their populations have been devastated by some kind of white fungus virus, or the "clean windshield" phenomenon associated with the devastating collapsing insect populations they probably depend on, and so it's not a big leap to think $100MM project is being mobilized in the face of a serious existential threat to their survival.
If the best you could think of is that "a" bat might "possibly" get injured it's a dramatic understatement of the kind of environmental threats they face. And you don't have to be anything more than a bit of a news junkie to know that.
You can, but you have to know where to look, and you have to have some idea of what you're doing. The benefit of Ollama is that the barrier to entry is really low, as long as you have the right hardware.
To me, one of the benefits of running a model locally is learning how all this stuff works, so Ollama never had any appeal. But most people just want stuff to work without putting in the effort to understand how it all fits together. Ollama meets that demand.
Very well done! This hearkens back to Asteroids, but it feels very novel.
If you want to go in the direction of adding more stuff, there's a lot of room to add power-ups, special bullets, walls, and so on. But this simple game is quite elegant as it is and doesn't really need any of that.
> Also, these assistants (for now) appear to exhibit no common sense about what is “much”, “little”, “exceptional”, “average”, etc. For example, after measuring a consumption of 3.5GB of memory (!!) for solving a 3-disk problem (due to a bug), the assistant declared all was well...
That describes a good portion of my coworkers.
reply