No, automation doesn't reduce jobs, i.e. doesn't reduce consumer spending, as consumer spending is determined by output, which automation boosts.
The savings from automation in a particular sector are spent elsewhere — wherever services are more costly (in labor). That's the dynamic behind Say's law, which shows that spending on less automatable jobs like barbers and physical therapists increases as automation reduces costs in other sectors of the economy.
I understand this is a well-developed economic theory and I am complete uninformed, but this doesn't make intuitive sense at all.
If 1 million prep cooks are replaced by robots, will food become cheap enough that those prep cooks can all get jobs as barbers, and the money people spend on food will shift to haircuts?
Will the food be so cheap that all those prep cooks can afford to learn to cut hair?
Also consider the money velocity of a human vs a robot. A human is probably paycheck to paycheck spending everything they earn. Robot earnings go back to company, which makes the stock go up, 90% of which is owned by billionaires who just keep hoarding and hoarding.
It takes real effort to maintain a solid understanding of the subject matter when using AI. That is the core takeaway of the study to me, and it lines up with something I have vaguely noticed over time. What makes this especially tricky is that the downside is very stealthy. You do not feel yourself learning less in the moment. Performance stays high, things feel easy, and nothing obviously breaks. So unless someone is actively monitoring their own understanding, it is very easy to drift into a state where you are producing decent-looking work without actually having a deep grasp of what you are doing. That is dangerous in the long run, because if you do not really understand a subject, it will limit the quality and range of work you can produce later. This means people need to be made explicitly aware of this effect, and individually they need to put real effort into checking whether they actually understand what they are producing when they use AI.
That said, I also think it is important to not get an overly negative takeaway from the study. Many of the findings are exactly what you would expect if AI is functioning as a form of cognitive augmentation. Over time, you externalize more of the work to the tool. That is not automatically a bad thing. Externalization is precisely why tools increase productivity. When you use AI, you can often get more done because you are spending less cognitive effort per unit of output.
And this gets to what I see as the study's main limitation. It compares different groups on a fixed unit of output, which implicitly assumes that AI users will produce the same amount of work as non-AI users. But that is not how AI is actually used in the real world. In practice, people often use AI to produce much more output, not the same output with less effort. If you hold output constant, of course the AI group will show lower cognitive engagement. A more realistic scenario is that AI users increase their output until their cognitive load is similar to before, just spread across more work. That dimension is not captured by the experimental design.
>472 different U.S. law enforcement agencies at the Federal, state, and local levels have the ability to directly query FinCEN's database of CTRs, suspicious activity reports, and more. This amounts to around 14,000 law enforcement officers who can search through the personal financial data of American citizens. In 2023, these 14,000 users conducted 2.3 million searches using FinCEN's query tool.
>FinCEN's data can also be downloaded in bulk form to the in-house servers of eleven different federal agencies, including the FBI, ICE, and the IRS. Bulk access (also known as Agency Integrated Access) means that the FBI, ICE, IRS, and eight other agencies don't need to use FinCEN's query tool. This bulk data can be access by another 35,000 agents. Alas, FinCEN doesn't track how many in-house searches were conducted by these agents in 2023, but I'd guess it's in the tens if not hundreds of millions.
Yes but in this case we have caught someone red-handed, why would you bring this other stuff up other than to distract from this known abuse? Sticking with topic at hand, do think we should prosecute those involved in this theft of data?
When private information is as widely shared as it currently is by government bodies, I think it's important to point out selective coverage of privacy violations, while the systemic encroachment of privacy rights continues unabated by 20 plus agencies.
In fact, a contrarian political entity created by an outside party — as the DOGE is/was — may be the kind of shock to major institutions that could lead to real positive change, in terms of greater transparency into and accountability over how these major institutions operate.
It would be interesting to see the list of past trees. The most famous I can think of Donar's Oak (also called Thor's Oak), which was revered by Germanic pagans, and felled by Saint Boniface.
>US on the other hand, has flatlined to the point where we think stuff like trans athletes in sports are a drastic enough reason to elect a president who is a convicted Felon.
This is very one-sided and unfair. The trans stuff is indicative of a larger social movement. For example, in the U.S., it would be illegal to use IQ tests to hire employees while in China, that's practiced. China is far more meritocratic. The U.S. is driven far more by ideology, and the trans stuff is an example of that.
And someone on the other side of the aisle would point to the prosecution of Donald Trump as politically motivated, where opponents found an obscure law that he violated and charged him with 34 counts based on the 34 forms he submitted with the expense mislabelling.
> China is far more meritocratic. The U.S. is driven far more by ideology, and the trans stuff is an example of that.
I'm guessing you never lived and worked in China before? People who get jobs because of guanxi are not rare, even today, and ideology is far more important in China than in the US, it is just that the ideology is very different from what people are used to in the states.
China definitely relies on ideology quite a bit, the difference is the government controls that ideology because they understand correctly that the people can't be trusted.
The Chinese government’s territorial claims in the South China Sea show near-total disregard for international law. China has constructed heavily militarized artificial islands roughly 200 kilometers from the Philippine coast — and more than 1,000 kilometers from the Chinese mainland — in order to assert control over waters that, under the UN Convention on the Law of the Sea and a binding 2016 ruling by an international arbitral tribunal, lie squarely within the Philippines’ exclusive economic zone. China lost the case on the merits and simply rejected the ruling.
I'm on the "outside" of this argument - never owned a gun yet and not in the US, but the right to life (not to be shot) can be exercised by protecting oneself from guns, with a gun.
Here we're discussing how attacks against privacy are totalitarian and how more and more governments are on their way to become totalitarian regimes, but we don't agree that people having guns is a good defense against a totalitarian government. We talk about police or ICE overreach, but don't talk about what would happen if that overreach expands even more.
That's kind of a jump. The 2a is cool, but gun deaths outpace car deaths now and 2a people refuse literally any of the protections we have against car deaths. Whereas a 15 year old jerking it to a pornstar hurts no one and these people want to completely ban the 4th amendment.
The great tragedy is that we already have a practically unlimited and environmentally safe source of energy, which is nuclear fission. And we simply don't use it at a significant scale because of irrational fears about meltdowns.
It is not rational aversion. Nuclear is currently 10,000 less dangerous per unit of energy produced than the largest sources of energy: coal, oil and natural gas. We could afford to let nuclear get 10x less safe, so that it becomes vastly less costly to deploy, and a very possible result would be that it would replace the largest sources of energy, and would still be three orders of magnitude less dangerous than the sources of energy it replaced.
Regulation is inescapable, because the maximum damage from a nuclear accident would exceed the value of the company operating the reactor. A rational business treats any liabilities larger that what it could pay as equivalent, regardless of how large they could become, and hence will underinvest in safety measures.
And I'm sure you will agree there is a great and sorry history of nuclear efforts failing to achieve their cost targets. At this point, it is clear that such targets are sales numbers, not something one should actually believe. One cannot make this history go away just by wishing, as nuclear advocates like yourself seem wont to do.
I agree fossil fuels should go, but that's not an argument they should be replaced by nuclear. It's the argument nuclear advocates used to be able to lie back and comfort themselves with, but then you all got blindsided by renewables and storage zooming past you. You have to address those now, not the old competition you wished you were still running against.
Of course regulation is necessary. My point is that current nuclear regulation is disproportionate to actual risk, and that this mismatch has made nuclear uncompetitive relative to energy sources that are demonstrably far more dangerous on a per-unit-of-energy basis.
Even compared to solar, nuclear has a stronger safety record when measured by deaths per TWh, and this is when taking into account the worst nuclear catastrophe, Chernobyl. I am not arguing that the future should be all nuclear, or even predominantly nuclear. I am arguing that the present regulatory regime reflects a mispricing of risk, particularly relative to hydrocarbons, and that this has pushed us into a suboptimal energy mix.
On cost overruns: the strongest correlation is with regulatory ratcheting, which also had harmful second order consequences for cost control from failing to reach larger scale construction, like bespoke designs and loss of construction continuity.
That’s not true. They are physically massive, incredibly complicated machines with all kinds of large scale pressure welding, forging, containment systems, 100s of miles of plumbing, and other serious large scale engineering. They will never be anywhere close to as cheap as something as dead simple & mass manufacturable as solar.
If they were intrinsically costly, they would have been costly in the U.S. in the 1960s, or in France in the 1970s-90s, or in South Korea today. It is because of regulatory ratcheting, and the effects of that (both direct and second order), that costs escalated.
NPPs are intrinsically Big Projects. The western world is almost universally suffering from Baumol’s cost disease - we cannot build Big Projects at a reasonable price anymore. Subways, bridges, NPPs, you name it - all cost many multiples of their inflation adjusted 1970 cost. And that’s before they inevitably blow their budget by 2-3x. Until you can somehow fix the labor / housing / management cost issues NPPs will not be affordable, even if you relax nuclear specific regs.
Mass manufactured things like solar and wind turbines do not suffer this.
Since the 1970s, large construction projects have been layered with several major regulatory regimes that didn’t exist when most affordable infrastructure was built. In the U.S., NEPA (1970) dramatically expanded pre-construction environmental review and litigation risk, causing planning timelines to extend by years. Around the same time, OSHA began rapidly increasing the degree to which workplace processes were regimented, which permanently increased labor-hours and limited on-site automation.
For nuclear specifically, this was compounded by post-Three Mile Island regulatory response. This increased the tendency to use bespoke designs, i.e. discouraged standardization, which prevented automation and the benefits of learning curves. That's why Baumol-style cost dynamics took over.
This same pattern shows up in subways and bridges. It’s not that 'big things can’t be built cheaply anymore", it’s that we changed the rules under which big things are built.
It's clear from the development trajectory that AGI is not what current AI development is leading to and I think that is a natural consequence of AGI not fitting the constraints imposed by business necessity. AGI would need to have levels of agency and self-motivation that are inconsistent with basic AI safety principles.
Instead, we're getting a clear division of labor where the most sensitive agentic behavior is reserved for humans and the AIs become a form of cognitive augmentation of the human agency. This was always the most likely outcome and the best we can hope for as it precludes dangerous types of AI from emerging.
The savings from automation in a particular sector are spent elsewhere — wherever services are more costly (in labor). That's the dynamic behind Say's law, which shows that spending on less automatable jobs like barbers and physical therapists increases as automation reduces costs in other sectors of the economy.
reply