blululu

Ok, this is kinda cool. In case you wanted to read the actual specs: https://www.hmd.com/en_in/hmd-touch-4g/specs?sku=1GE001CPG30... and https://en.wikipedia.org/wiki/HMD_Touch_4G RTOS Touch on a Unisoc T127... that's cool. Amazed they can get the price that low.

This all sounds like a lot of backpedaling and “well actually” kind of stuff.

“Stochastic parrot got picked up and interpreted by other people as a minimization or an insult. It was not meant that way. Other people might be using it that way but that’s not how I intended it”.

Yeah that’s because it was chosen to be an insulting phrase.. Parroting is only ever used as a pejorative phrase. But sure, everyone else mindlessly parroting this line is the problem here.

This paper was always lousy, but it has really not aged well. We are living in a world when where an LLM has solved an Erdos problem. In a world where LLMs produce novel results that rival human thinking any conceptual reduction of an LLM is going to start inviting some unpleasant comparisons with human thinking.

Yeah. I'd agree. This is a pretty quick and easy law with very few downsides. There is also a "good house keeping" aspect to this. The habit of keeping a house clean and tidy makes it easier to do the bigger and more substantive improvements. I'm a bit surprised that New York Times is reporting on such a small law passed so far away or that there are such negative comments about such laws here, but such is life.

This is a pretty nothing law when you look into it.

All it does is standardize "sell by" labels to be more transparent. The extent of the ban is the explicit phrasing "Sell By" which is itself confusing since the manufacturer has some estimated consumption interval that is tacked onto their actual expiration date. Sell By gives the customer has no idea how much padding they are adding to their estimate and when the thing actually goes bad.

All this law really does is it standardizes the labels manufacturers can use to "BEST if Used by" for quality concerns or "USE by" for safety concerns. A lot of manufacturers already do this, so it is a pretty minimal law There are probably more pressing issues in Sacramento, but a small improvement is always welcome.

That’s a fair framing. I like the idea that the two debates could be separated. Personally I find it hard to believe that a super intelligence would not completely disempower humans over any reasonable time frame (50 years), but you are right that there is a chance. There are some optimistic scenarios either way but the odds are indeed long. You can approximate things reasonably well by framing the priority as whether you think P(take off can be stopped) is higher than P(survival post take off).

This could be true, but I think it amounts to wishful thinking. The human empowerment phase of AI is transitory and brief. Once you get into RSI and AGI the human is increasingly left out and disenfranchised. The multiplier of your personal effectiveness directly proportional to how much the AI is doing relative to you. Being 2x more effective is great - AI does half the work. 5x, 10x, 100x? When the AI does 99% of the work you maybe start to worry - the bottleneck to 'your' productivity is now you and you are going up against much faster and more powerful entities than yourself. This mindset is rooted in a short term gain that risks a long term loss.

Feel like there ought to be some Sinn Fein backed point to point connection in Belfast that leverages the Good Friday Agreement to get around this.

You can use an llm to get out of doing homework but you can also use it to ask every question you would ever wanted in a 1-1 tutoring session. The problem is kids will use it to cheat on their homework. If we can’t deal with that problem then a ban is necessary. But these things can be phenomenal teachers if you use them properly.

The author’s claim that Claude is a multiplier for skill is probably true for now but it also feels like cope inspired by usability issues with Claude. The advice is all good, but none of it is especially clever or impressive or hard to grasp. The multiplier just comes from the fact that anthropic hadn’t taken this essay and several similar ones and incorporated their feedback into the product. This is a pretty shallow most of expertise that anthropic ought to automate in a week.

The two points are not contradictory. And historically playing the boom and bust game has blown every player out of the game. This time could be different in so far as Chinese firms can run insane losses of deemed vital to national interests. This is a problem for these firms, they need to balance that with the existential remains risk that over production can drive them bankrupt.

Just piling in to this because it needs to be stated with emphasis. Wikipedia foundation is not Wikipedia. Their donation campaigns are highly deceptive. Pennies on the dollar will go to supporting and maintaining the actual encyclopedia. WMF is as much a liability to the encyclopedia as a beneficiary.

IDK man - this is sort of true, but I think you under-estimate how quality and price scale. A Jumbo-Choco-Monaka at 7/11 is still a fantastic value at ¥160 even if you adjust for purchasing power. GDP Per Capita (PPP) is about $85K in the US and about $60K in Japan, but even granting a 2x increase for California then a $2 choco-monaka would be a steal. As it is, I just spent $4.50 for an Its-It about an hour ago and while I am quite a dedicated fan of these things I would have gladly forked over ¥700 for a Chocomonaka if such things existed in California. I realize that people don't live out of 7/11 for their daily groceries and your point has some validity, but the quality/cost is still a great deal relative to what you would get in America.

Looking at the history of the memory industry the biggest risk is that a firm would over produce and go bankrupt. Maybe this time is different but so far no memory chip maker has gone under because their competition increased capacity.

Québécois separatism is also driven by a single party with no plan for what to do with all the other groups. I also don’t think that an independent Quebec would be a good idea, but they have leveraged the idea to get equalization payments and increased voting rights. These concessions largely come at the expense of Alberta, so it shouldn’t be hard to see why people would be frustrated without any cia operations.

Yeah, total us payroll is around 15T. They are basically spending $10 per employee per year. Or 1/10000 of the total spending on wages. This is actually small in terms of political spending. There are random scam ballot measures in CA that get more spending per voter than this. *edit: the actual report details this is just for services related to efforts targeting their own work force via lawyers and consultants. The total spending on the issue is likely much higher: https://www.epi.org/publication/u-s-employers-spend-more-tha...

Assuming that this is a good faith response and not merely a bot (and I have very little reason to believe this given your history of spewing AI slop): I think this is a lot of bullshit. You're either lying to me or you're lying to yourself.

>Yes, our experiments get attention, but I wouldn't call them publicity stunts.

It sure looks a lot like your startup's publicity stunt. There is nothing wrong with marketing a product that people want.

>The point is to give the world more data points of what happens when you put AI out in the world and let democracy do its thing.

How generous. I suppose you can justify any poor behavior as raising awareness about the consequences of said behavior. Littering trash to raise awareness of pollution? Even oil companies don't try to pull such a line on pollution. The public opinion on this stuff is pretty decisively negative: the institutions just haven't caught up yet to make fines for it but I don't see any reason to force an acceleration here. (more useful applications of AI will experience a blowback from the more anti-social applications).

>Soon, a lot more people will do this at large scale because it will be easy. I hope we decide where we want AI in society before that.

We're all looking forward to this future. Other people doing something is not an excuse to do it; especially since they haven't even done it yet. From the look of things you have very clearly made your choice about how you want to use AI in society. I would point out that you're not particularly hard up for cash. You clearly have lots of talent and ability. You could be putting this to a better use. I would be happy to offer you a job. You really don't need to be doing this.

>Personally, I'm very much pro a pause on large AI training for example. I hope our data could be useful as a grounding in such discussions.

This would be a more convincing line if you weren't actively trying to profit off using AI for the destruction of the commons. Using AI to cure cancer or male pattern baldness gives us something new that we can be excited about. This just gives us something crappier than what we had at someone else's expense. Putting people out of work with AI is going to cause problems. Maybe it's inevitable and maybe it can be good in other ways, but I strongly doubt it is the path to minimize p(doom). If you believe in a pause, then simply stop. Yes you will lose money: ask me how I know. What hope does a Pause have if smart and talented people are all so excited to be blitzing to an undesirable Nash equilibrium? You don't need to do this. You know it's not the right thing to do and that people don't like it. You don't need to run an experiment to know this, but you have now run three with the predictable outcome. There really isn't much excuse left. Please focus your talents and abilities on something better. AI can do some many things that are simply impossible today. We really could achieve new heights but this project really doesn't feel like the dawning of a golden age.

This is their third publicity stunt in the past couple of months. It follows the exact same pattern of attention seeking at the expense of the commons. At this point they seem like a bunch of low empathy jerks. They are gleefully describing their progress in developing yet new frontiers in AI slop. I’m sure they are all very pleased to think that they will be profiting from a future where ai slop is everywhere. I could go on but it’s tedious.

Not really. Public school teachers are well paid when you consider pension, tenure and low stability. Working in industry can pay more than teaching but this is not a guarantee. A public school teacher doesn’t get fired when they turn 40, or laid off in a downturn. It’s less about money and more about qualifications. having a chemistry degree is not the correct qualification to teach chemistry in a public high school. The labor market between stem practitioners and stem teachers are really not substitutes in any practical sense.

To be clear the shoplifters in question are all rich themselves and stealing expensive items they don’t need. The original article is about students at one of the richest and most prestigious institutions in the country. None of the criminals are poor by any stretch of the imagination. They are just lousy people who are smart enough and entitled enough to try to justify their bad behavior.

Out of curiosity how much of this is a manifestation of the utility of LLMs? I get the current political impetus right now but also the barrier for swapping out an infra stack was also much higher 2 years ago. From my own projects major swaps are now relatively trivial which means that vendor lock in is weak.

I'm sure there are plenty of caveats and breaking points, but if we do adhere to the claim that an LLM coding tool is a nondeterministic sort of compiler then it really does make sense to pick the most performant language available. Obviously there are caveats of libraries and native advantages of various languages. I've been doing stuff in C++ for the past month or so and the only slow down from the language choice is compilation time.

Those McKinsey/HBR studies are trash. They privilege the hypothesis, overlook the obvious ecological fallacy at play and add in a bit of a sampling bias for good measure. The fact that East Asian Economies are all booming and exporting globally with ~0 diversity and unique cultures ought to refute this notion. I'm sure there is some no true scotsman line you can play here about how the true meaning of DEI, and I would agree that the stated goals of DEI are all laudable. But in practice these initiatives often amounted to unprincipled discrimination and venal power grabs, which is why they are so widely despised.

I go there routinely to inspect the sealions. They always cheer me up after a rough day.

Sea otters and River Otters are quite different species and they live in very different habitats. My understanding is that Sea Otters don't really travel very far. They have serious predator and there are large gaps in kelp Forrests so they ~stick a kelp bed that they like and live there. River otters in contrast are ~the apex predator of a river and have a pretty extensive range. A river otter traveling 100 miles is not uncommon.

Sadly the sea otter populations were decimated by Russian and later American hunters in the 19th century and never really recovered. My understanding (limited to California) is that the southern sea otter was thought to be extinct until they found a population in Big Sur. They have since been slowly trying to expand its range north, but 80 years on and they have not gone much farther than the Northern end of the Monterrey bay. I suspect that with enough time and patience the Sea Otter will slowly be reintroduced along the entire coast, but it will take a while.

Finally some good news. I had no idea the Ohio Department of Wildlife had such an outstanding role in the recovery in the Great Lakes region. River otters are really interesting creatures.

Humboldt State supports an active citizen science project to observe and document river otter populations in the ~watershed of Humboldt Bay in Northern California. It's a cool model for engaging with the community to help protecting its natural resources. I could easily imagine similar resources in other regions. * https://hsu.reclaim.hosting/NorthCoastOtters/

They should make finder’s search function actually work so that you can open apps on the latest release or macOS. Everything else is irrelevant when you are shipping massive UX regressions and bugs.

It’s a made up toy problem. It exists for fun. The stated problem has some implicit assumptions. But you can rejigger the rules and assumptions to tweak the incentives and ethics. That’s the whole point. You could take the puzzle and apply it to a band of pirates held in a jail. That might make the outcome more obvious. Or you could imagine what would happen if the voting order were sequential. These are all just different formalisms that are fun to speculate over, but the rules can be interpreted many ways.

Good points, though I think cooperation benefits the ethical outcomes for both sides.

If we all work together to make sure that as many people press the red button as possible, then we can minimize the damage. The problem with the blue campaign is that the outcome gets progressively worse until it gets to the best outcome. 49% mortality is high and terrible unless you are very sure that the red campaign is going to lose. The ethical take on the red side is to minimize blue votes to zero.

What if you entertain the variant of the question where a percentage of red votes die in the event of a blue win? It makes pressing red less advantageous, but also it totally changes the moral balance depending on the percentage.

The first point is interesting. You could fork the question over this and have a few variants:

1.) The pure form where the button presses and restricted to legal agents (i.e. people with credible legal standing over their choices). 2.) The mixed form with the caveat listed here inclusive of all humans whether they are even physically capable of pushing a button. 3.) you could also go for a more expansive scenario that takes 2 to the extreme and includes animals as well.

1.) gets to the game theoretic form of the question. 2 muddies things, and 3 sets up a case for blue since the non agentic voters asymptote to 50-50 and a slim edge is morally preferable to killing half.