Whether it's high altitude drone swarms, terminally guided artillery munitions, hypersonic rail guns, or high energy laser defense, all are orders of magnitude cheaper than the interceptors and could be less than the cost of the (nuclear?) missile. It's true that generically defending against nukes is basically a fools errand, but if they're (also stupidly) limited to putting them on ICBMs with non-detonating fail safes, then it's probably economically doable and cheaper than the $10T forever war.
I'm sorry, the whole framing of this (OP) question/answer seems artificial and fundamentally silly.
It looks like a lot of them are missing something big. I'd think the two big ones are the evaporative cooling as you pour into the cup, and heating up the cup (by convection) itself. The convective cooling to the air is tertiary, but important (and conduction of the mug to the table probably isn't completely negligible). If there's only one exponential, they're definitely doing something wrong.
I'd like to see a sensitivity study to see how much those terms would need to be changed to match within a few %. Exponentials are really tweaky!
It's a mix of course, but I think it should be mainly that and evaporative cooling. Evap is _very_ effective but will fall off rapidly as you get away from boiling. The conduction into the mug will depend a lot on the mug material but will slow down a lot as the mug approaches the water temperature.
I'd be very interested in seeing separate graphs for each major component and how they add up to the total. Even asking the LLMs to separate it out might improve some of their results, would be interesting to try that too.
Yes, since they didn't explicitly list the evaporative cooling when the coffee was poured into the cup, I suspect it was not included (as if the coffee started in the cup). That means that the starting temperature is off and screws up all the other calculations.
The evaporative cooling as you pour into the cup is when the coffee is at the highest temperature and has the most surface area even though it only takes a few seconds. One could test this either by including it explicitly in the requested calculation, or by putting the fill spout directly at the bottom of the cup when filling.
I agree and disagree with different parts of this article. I've read/seen a lot of the source documentation so I think there's plenty of hyperbole, even if there's a nugget of truth.
Because the “The AI Bubble Is 17 Times the Size of the Dot-Com Frenzy — and
Four Times the Subprime Bubble” (oh, and also, there is also a new subprime
bubble—and it’s already collapsing, which will make all of this worse).
It's not 17 times the size of the Dot-Com, certainly not scaled with the total market valuation. Much of the money that has been "promised" has not be delivered and there is a LOT of circularity to the funding, but it's not $Trillons yet. Many of the companies are still private, haven't gone up much, or are only fractionally floated so the numbers look big. But, it doesn't look like there's a huge moat, and it's going to be expensive to pay for those training servers with inference. The depreciation is all wrong, they're not in the right places, and power consumption isn't optimized. TSMC is probably gonna sell more chips to make it so.
At the same time, it's great to be a user of a "free" product. LLMs work as well as Google search used to! It's great! You can't believe everything you read on the internet, but if you know enough to verify it, it's incredibly useful. If you build an OpenClaw footgun, you deserve the consequences, even if your other victims probably don't. Will the "AI" companies end up paying for it all by exfiltrating their "customers' data? Facebook and Google did.
Can't speak to the OP, but lots of technical work (and frankly many trades are also technical) doesn't lend itself to text based documentation and teaching. Software, translation, non/fiction writing (like marketing and sales) all do. I think LLMs will take a significant part of those businesses, because I don't believe there is a Devon's Paradox for code -Tractors- Agents.
At the same time medicine, hardware design, good industrial, and specific domain knowledge (problems you solve in assembly or control loops) that are fundamentally proprietary and aren't well documented will continue to have value even when LLMs make solving the problems around them easier. Those might have increased leverage, at least for this round of LLMs. Now, maybe they succeed in World Models, but that is not today.
Really, I don't know what "kids these days" are going to do. I couldn't have predicted the influencer boom 15 years ago, but I also think there are geopolitical risks that are probably bigger than that shift, and "synergized" with the push to AI Everything, it doesn't look like a good time to be a learning/working human.
I suspect that was deliberate. DoW is the preferred nomenclature but DoD is still technically correct.
The article is phrased in a way to imply that the author would rather the publication maintain independence. It is probably the last time she will be permitted to say "department of defense".
You're implying that insurance companies will allow prices to fall and lower their profits. That seems like a really unlikely event in the current economy. They fire a lot of doctors and nurses, but they won't lower prices.
The ACA requires 80-85% of health insurance to go toward medical care (medical loss ratio). The way they work around that is to figure out how to charge more for medical care.
I'll offer a hopeful rejoinder. Perhaps, when AISlop generates enough of the same old story "guaranteed" hits for the mass market (and book covers to go with same), the editors will switch back to something that is novel and unlikely to be generated.
Think about what happens when you feed the first few books of a series into long context llm, along with their audience interests, pitch lines, plot summaries and character guides. When each element is multi-shot rather than zero-shot.
Yep! Both private and public insurance funds are nickel and diming wherever possible to make up for losses during the COVID pandemic which went on for 3 years.
Edit: can't reply
> This started decades before COVID, and has been summarily ignored for the entire time. We’re still insuring properties in areas we known sea level rise and climate change make uninsurable
It's not property insurance that's causing the issue as I as well as the article pointed out.
This started decades before COVID, and has been summarily ignored for the entire time. We’re still insuring properties in areas we known sea level rise and climate change make uninsurable, we still refuse to let the government negotiate with or dictate rates, we still tie healthcare to employment, and we still let manufacturers sell top-tier/unrepairable vehicles while constantly discontinuing anything remotely affordable or repairable.
The market isn’t going to correct itself now that we have enough data to know where to squeeze next, and how hard it can go. The only way to unwind this disaster is meaningful regulations and wholesale reforms.
How much has the sea level actually risen? Our most expensive flood in terms of insurance losses was the Great Flood of 1993 which had nothing to do with sea levels.
Here in the Pacific Northwest, the catch-and-release of car thieves drove up costs for everyone. If I change my address away from the craziness, my policy drops by half.
I have always only purchased maximum liability only insurance, and my premium went from $40 per vehicle per month to $50 per vehicle per month over the last decade. In Washington.
Rethinking Attention-Model Explainability through Faithfulness Violation Test Yibing Liu, Haoliang Li, Yangyang Guo, Chenqi Kong, Jing Li, Shiqi Wang
https://proceedings.mlr.press/v162/liu22i.html
https://icml.cc/virtual/2022/spotlight/18082
reply