CharonY

Not according to the president.

If costs can be externalized it is considered a smart business decision. Regulatory framework should limit that, but often doesn't, or does it too late. Well, there is an argument that additional LNG generators could bolster existing energy infrastructure and because of the huge capacities of data centres to absorb excess and regulate demand it can be a net boon, as you mentioned. A bit of an issue here is the significant added consumption (and emissions) that this causes. I think one issue is that there is a big push to get even more of the paste out and it is not entirely clear to me that it makes sense. There is an urgency that almost seems to be designed to throw caution into the wind. Economically it seems to be precarious at best, and at least based on my reading it seems that the bubble will have to pop eventually. It is reminiscent of the dot com bubble. I may be wrong, but I don't think that that bubble had also such ecological and physical infrastructure impact as the current situation. The main argument I have seen for the need for massive investment now relates to national security. But even if being just slightly ahead would create an advantage in protecting and/or initiating cyber attacks at scale, the companies are not under direct control of the respective governments (other than in autocratic countries, including China). But even with some of the regulations in place it does seem that there is going to be a lot of trust put into those companies to secure cybersafety. And of course, if USA wins the race (whatever it means) it could also spell disaster for its enemies, such as apparently Greenland and Canada. So following that logic, the massive investment are basically a race to weaponization, like in the early days of nuclear ambitions. Only this time not in a nationally coordinate way but via massive investment into, I would say dubious characters. I think we really need to design a better toothpaste dispenser.

That is very good to know. So it is about optimizing the margins? I assume as everyone and their dog wants to be hyperscalers it is getting competitive?

Well, some do require closed-loop systems, though I guess not so much in desert areas. Issue is higher cost, lower efficiency, especially in, say, deserts. An additional issues seems to be that coolants used in water-free systems are harder maintain and can lead to more toxic leaks. Usually that is a problem when data cenres close and someone has to clean them up. One argument I heard is that that this gives them more flexibility to address fluctuating power demands (for the given cost that is). Which is why the claim that AI is going to solve our energy and climate change problem doubly ironic.

It depends on what parameters we are looking at. I think different health systems might include different parameters in a standard blood test. Those most sensitive to diet often are conducted after fasting. For fasting blood glucose level you are typically asked to stop eating for 8-12 hours before the draw, for example. But most I can think of (e.g. blood cell counts, lipid profile, TSH, total protein etc.) generally are little impact from these types of short-term shock and/or normalize within hours.

Of course, I was referring to some of the challenges that social workers have brought when talking about reaching out to unhoused folks and the diversity of challenges they face. That being said, these initiatives are notoriously underfunded to begin with, and many, if not most help organizations focus on immediate harm prevention more than anything.

I would also like to add something that has been implicit in some of the posts. The solutions are not universal. The issues leading to homelessness for example are extreme diverse and so are potential solutions (if they are known at all). In some cases, especially folks that are at risk to losing their home due to loss of job can be assisted with bridge funding, retraining and other measures. However, folks that have lost their home for an extended time, often require additional assistance benefit from such funding. That part touches on it, and there are a wide range of reasons why some refuse such accommodations, ranging from experience of violence to deep mistrust to organizations and so on. Then add to this health issues of various sorts, which requires bespoke, and thus expensive support to really help them.

Oral tradition is notorious for misinterpretations even within generations. Cultures that depended on oral tradition often had fairly elaborate systems to reduce that. While books can have similar issues, the baseline is less volatile short-term, at least.

If you are talking real transfer rather than training on a specific subset of skills, I disagree. Or rather,I think for anything somewhat complex you need both. Books for the broad foundation, and direct guidance to fill in gaps and consolidate the ideas. Fully oral training is hard. I have talked with folks regarding traditional knowledge transfer in oral societies and typically there are dedicated positions (knowledge keepers) who are heavily trained to retain and transfer information. Once trained, the person is highly capable, and it could be considered effective, but the effort needed is much higher than textbook assisted knowledge transfer.

The US is doing a lot of silly things but this one is hilarious (and dangerous) on so many levels: https://thehill.com/policy/defense/5969935-hegseth-testosterone-testing-dod/ There are so many layers to it. This is biologically and medically dubious, and shows what happens if lack you weaponize stupidity to max lethality or something. It also shows that the weird one-dimensional gender/sex view of these folks is not just a slogan, they really believe in it. (I.e. testosterone is manly, more manly -> more lethal). And on top it betrays a rather limited perspective on modern warfare (how much testosterone does a drone operator need?). And then also the absolute hypocrisy regarding hormone treatment of personnel. Added to that, we get a view what happens if influencers actually run things (see RFK for the the show on the health side of things). It is like a multi-layered idiot sandwich with misinformation sauce.

The concern that folks opt not to learn if an answer is already available is a valid one and we do issues with that in practice. However, regarding teaching, it can be a good supplement when done correctly. I mentioned that I think that there are certain elements to teaching that seem to be used for a long time, such as dialogue. And in areas where the knowledge framework is well-defined, chatbots are now good enough to simulate such dialogues. Where it can serve a role is then to supplement teaching. E.g. provide a way for students to fill in the gaps in their understanding. An example if they miss out or misunderstand a step in a proof or complex chemical reaction. But the difference here is that the student has to use it to develop their thinking rather than demand answers. A human instructor would try to give hints without outright giving the answer, and I assume that it should be possible to train a chatbot to emulate that.

I think it is just more exciting for male pupils. And at least to a certain degree it can also be framed in a way that is less uncomfortable. I know two highschool teachers who are on the opposite spectrum of this. One is way into the minutia of battles and military hardware and can for example, talk at length about the strengths and weaknesses of the Spitfire vs the Bf109. And I learned way too much from it about that. The other is predominantly interested in social aspects, including how folks reckon with their past. One is way more popular than the other with pupils.

There is increasing evidence that replacing textbooks with computers and tablets might have degraded teaching effectiveness. Yeah, well an issue is that many of us are (by necessity) ultra-specialized and it can be a bit tricky to think broader in some cases. I found that there is a huge variety in how history gets taught. The best teacher in highschool I had, focused heavily on social context over individual drivers, which I found was a much expansive view that made you think more. At least, compared to troop movement and focusing on individuals.

I think it depends on what you focus on. Topic-wise cross-fertilization can be very forced. But the part that you can emphasize are transferable skills. Careful reading and interpretation can be applied to both, poetry and natural science paper. It is the framework of interpretation that has to change. And highlighting those differences can crystallize the skills you need to do either. And there are topics where there are overlaps. For example, as a natural scientist, if you work even a little bit with humans, certain principles and ideas on community engagement are incredibly valuable to make inroads in your research. I think there is some value in the idea of rounded training and education, but one need to put some thought behind it.

To a certain degree, there might be. As a whole I found a combination of dialogue and hands-on work to be the best way to convey concepts in many domains. Regardless of all the add-ons that companies, administration and even some areas of research suggest, none of the stuff I have tried was giving better long-term results. It is, however, not easily scalable to very large groups.

Depends, CGI traditionally required humans to generate the effects. AI sort of interprets the input but creates the outputs based on directions. I.e. the human input is vastly reduced and, in theory, could be eliminated entirely.

I am still not sure about that. The argument for AGI was that it will be able to improve itself almost infinitely. Perhaps this is a separate discussion, but I am unclear regarding limits and boundary conditions. There are, for example, already now hardware and energy limits. As the system works entirely different from how our (biological) mind works, I am not sure where the respective limits and differences are. But I think beyond that, our abilities to learn has already been changed (I would say degraded) by other things (including cellphones as portable dopamine delivery systems) and for a long time we really haven't found a teaching system that can address that.

My thinking has slightly shifted on that matter. Originally I was in the camp of teaching how to think rather than feed them info. But along the way it seemed that folks became incapable a) to find any information, despite increasing ease and b) have no idea how to handle information and c) do not benefit from training in problem solving as they used to. I think that at least some level of memorization and other boring routine exercises actually somehow strengthen mental capacity to due the trickier parts. And perhaps more importantly I think (I have no experiments to support it), it does counter the ubiquitous deterioration of attention span a bit. I am not joking when I say that I have students who by the end of the sentence do not remember the start of it anymore.

No doubt and higher education in general was traditionally geared towards understanding, not repetition. The commercialization of the process has made it harder to maintain. Depends a little bit. In the Chinese system there is still a lot of emphasis on memorization, yet they fare quite a bit better than their North American counterparts. One of the reason I think is the effort put in. I.e. emphasis of effort over efficiency. The former has the chance to train other cognitive abilities. The latter leaves more time for TikTok.

If there is access to LLMs somehow (say, smart glasses) then the answer is no. The other issue is that over the last decade or so, we typically always use a mix of memorization question to deeper question. An easy way to figure that out is to make an oral exam based on what they present, for example and try to figure out why they arrived at certain conclusions. The issue was always that (obviously) fewer get the reasoning part right, but even if folks fail, as long as they build a foundation (probed in part by memorization) over time they still can develop expertise to figure things out. I.e., it is a way to give students time to develop their thinking muscles without immediately failing them. As such, in higher classes the why questions traditionally were a higher proportion of the quizzes. Now, the issue, especially across NA is that high school (and probably already earlier) students learn exclusively what to put on a test without understanding the why. In fact, in their mind, the why is superfluous as it just adds baggage, after all the only thing you need is the right answer. And as there is a right answer for everything quite a few seem to think that trying to deduce it is wrong, you just need to know or find the answer (and weird as it is, before LLMs quite a few were turning to TikToks for answers). So, they do not even lack the skills, but even the awareness that reasoning skills are relevant. And they do get anxious and even angry when made aware as they feel the instructors are wasting the student's time by outlining the necessary background information. A common demand is: "just give us the correct answer and stop wasting our time." I.e. there is a lot going wrong especially in NA due issues with the education system (e.g. linked to tuition-based higher education), which, however, is part of a roughly global trend that indicates a general loss of reading, comprehension and overall competence. Papers are start to come out on this topic and they are a rather depressing read.

No, absolutely. However, in some forms of cheating (say, cheat sheets) they actually have to do some work. Being given the answer bypasses anything that could accidentally make you learn. I am more worried that not even starting to think but immediate turning to LLMs is increasingly becoming the norm.

Fair point. I just wanted to add that it is a bit scary how quickly also (or perhaps especially) young folks abandon their own learning and submit to the perceived superiority of AI answers. It has been a quite a few times when students were arguing about something and it turns out they were referencing ChatGPT, often not realizing that it runs counter to something that they were supposed to have learned. I found that whatever chat bot they are using, it is often hooked on what might be propagated as conventional wisdom with way more certainty that the lit actually allows. It is still very bad at figuring out biological nuances, but unfortunately we can already see how it erodes student abilities.

Apparently I have no idea how Langobardic art looks like.

My thinking goes a bit into a different direction, and I might be totally off. However, at least in my field, insights are rarely gained by extreme creative thinking about a problem. Instead, a lot is empirical, based on experiments you conduct, which provides unexpected outcomes. I.e. major breakthroughs happened often serendipitously, and are often based on new experiences. Now, once we give the systems a body to conduct experiments, it may mimic that. But as I see experiments and experiences as a limiting factor I am not sure how much robotic plus novel AI can accelerate discovery, say over an army of grad students, especially if the latter have access to robotic pipetting stations and other forms of automation. I am old, so I may have been repeating that though many times over already, but in an area such as biology, where the unknown far outweighs the known, ideas are less relevant than the resources to try out things. Edit: there are areas of exception in which computational approaches are are already heavily used, e.g. for big data analyses, and here the role of ML and other models are going to be highly relevant, of course. I think what I trying to express is how important experiences with reality (experiments, field observations, etc.) are for gaining novel insights.

I also saw a pre-print somewhere which suggested that some a significant proportion of information generated by AI chatbot originated from sources like Reddit. Findings such as these do make me question the claim that somehow an AI will be able to answer all questions we have.