I'm old enough to remember the computer bulletin board system (BBS) days fondly. I ran my own BBS for over a decade, in one form or another. At the end, I moved it to kind of a hybrid setup where I had 4 dial-in phone numbers for modems, but also a 128K ISDN circuit to the Internet. Most callers used it as a way to access some Internet resources while others started connecting to the BBS itself via the Internet.

Ultimately, BBS sysops (the jargon for system operators) realized we were at the end of an era. The Internet simply made the hobby of BBSing obsolete. Why limit oneself to using a specific bulletin board system and whatever limited options it has on its menus, when a raw Internet connection allowed hopping all over the world to content of all types -- even accessing multiple screens of it simultaneously?

Even back then though? I think many of us mourned the lack of personal connection that caused. On a BBS you used regularly, you had typically 100-500 or so somewhat regular users to interact with. Most were in your same geographic location. It wasn't uncommon to organize gatherings so people could meet the real humans behind the screen names/handles.Message topics were more relevant to where you lived. People could actually request 1 on one chats with the sysops, and most people got to know them pretty well.

The Internet doesn't "need to die", but maybe more people should make the effort to seek out and use web sites offering quality content, vs a habit of spending most of their time on big-name social media sites? I know for me, a good example for this is cars. I've always been kind of an auto enthusiast, and I used to always seek out a web message forum for any vehicle I purchased, and immediately join as as new member there. These days, I see all these enthusiast type groups popping up on Facebook -- but none of them offer the depth of information the dedicated web forums have. You just wind up with thousands of people in the Facebook group asking the same questions and wanting to show off photos of whatever they did cosmetically to their ride. The message forums always seem to have a few regulars who REALLY know the car inside and out. I'm talking master mechanics who have taken the whole vehicle apart and put it back together again, in their own garage. And they post a LOT of "how to" type message threads and more.

IMO, people are making too big a deal out of the touchscreen functionality. It's not really a big reason someone would buy this new Macbook Pro. It's simply something Apple said they were finally able to add to a good OLED display without adding weight or thickness to the machine.

I had a Lenovo ThinkBook 14" laptop my workplace issued me that had a touchscreen display, and I basically just forgot it existed on it. That said? It was finally useful in one situation where I had it on a docking station with dual external displays. I configured our VoIP softphone to use the laptop's own display so I could tap to call people from the directory.

This is just the latest, greatest iteration of Maxbook Pro with the latest ARM processor in it and a number of small design improvements. (EG. It has an audio profile now for listening to music with the lid closed on it. That's a feature I'd actually use sometimes and wondered why none of my other laptops ever had it.)

I have to think a bit to organize why and advantages and disadvantages of this, so I thought I'd share.

I define "tokenizing shares" as using a distributed blockchain with universal ability to transact as a way to trade shares.

Advantages of tokenizing shares:

* Universal access to fractional trading, not just inside a single trading platform. (Most likely, anyway.)
* Near zero transaction costs outside the spread. There are trading platforms that offer $0 commissions for regular retail customers, but there's still regulatory fees and these could potentially be skipped. (NOTE: There is likely a larger spread on a distributed ledger than a regular exchange, due to slower transactions. See below.)
* Ability to trade 24h
* Access to purchasers that don't want to or cannot use traditional platforms (e.g, crypto fans, people in developing countries). This may be stopped by "Permissioned ledgers", where there contracts that stop the transfer of stocks into a wallet unless it's on a whitelist.
* Resistance to total failure (due to the distributed nature of the setup)

Changes with good and bad sides:

* Slower trades; expect max speed of milliseconds rather than 1us
* Thus, a loss of the ability to do HFT (High Frequency Trading), which means that we lose the information transmission and market making ability of HFTs, but we also lose their ability to do front-running and decrease their ability to cause flash crashes and similar
* Loss of ability to shut down the trading. This remove an abusable tool of power, but it also lose a tool used to stop bad situation (like flash crashes), which is used to deal with bad market situations (like flash crashes due to trading software bugs), to avoid insider trading when major news is about to hit, and to deal with regulatory concerns.
* All trades are final. In the regular market, trades can be reversed by the courts or banks if done quickly (clearing in the US is in one business day, in Europe it's mostly two business days); in an anonymous, immediate clearing system based on a final distributed ledger, trades can't be reversed. This removes some types of risk, but adds in risks of theft (as we've seen with cryptocurrency) and makes mistakes more difficult to reverse.

Disadvantage if both on a regular exchange and a blockchain:

* The slower trades on the blockchain will lead to latency arbitrage. To avoid losses from price sniping, the blockchain market has to have wider spreads. Market makers will also typically provide limited liquidity in the slower market, to minimize the amount of loss the times they do get sniped.

Disadvantages of stocks on a blockchain in general:

* The distributed nature of the blockchain will allow frontrunning: A retail investor place an order in a market, and a frontrunning company with faster routing setup will then buy the stock elsewhere at a lower price, and re-sell it to the investor at the higher price. If there are distributed validators for transaction order, frontrunners may instead use techniques to bribe the validators (e.g, they may run validators themselves, and have their purchase jump the queue). The net result of both is that instead of negative spread going as profit to the retail investors, it goes to the frontrunners.
* It provides larger liquidity pools for criminals and terrorist organizations to do money laundering; e.g. Russia may be able to move billions around outside sanctions (though there are protocols in place that try to stop this)

Disadvantages for this particular blockchain at this time, for the companies that are resistant to getting on it:

* It is cheap to say no now, and they can't say no later
* There is little time to evaluate
* There is likely some level of legal risk

That's a quick attempt to sort my thoughts - they're not strongly researched for this particular blockchain, though I did a couple of searches to fill in details that I wasn't certain about.

That's nowhere near true. The current system is administered by regulated entities that have a decent amount of information about most owners. Are there ways to play the system? Yes. But it's nowhere near the same level of anonymity and flexibility as making stocks into tokens that can be freely transferred to anonymous accounts.

I would assume places felt unsure they'd really get their data back by paying the scammers who locked their files down in the first place.

There are plenty of stories about people who paid the ransoms and never got their data unlocked afterwards.

If someone came along, pretending to be the good guy who could break the lock for you or your business? Wouldn't you prefer to pay THEM, rather than rewarding the criminal who created the problem in the first place?

It's at least notable that the plastics industry documented the possibility that microplastics were accumulating in the human body. But with there being such limited knowledge on what harm they do, if any, just by being there? I don't see why this would have sounded any alarms back as far as 1974?

Comparing this to cigarette manufacturers is a little disingenuous, IMO. That's an industry that encountered mounting evidence that smoking was doing harm and even killing people, yet they doubled-down, paying off researchers to try to find as much contrary evidence as possible to deny it.

It seems pretty logical that ANY substance we're exposed to can enter the human body, and many of those substances could stick around in a microscopic scale. The more important question is whether or not it matters from a health perspective. We developed some of our plastics BECAUSE they were inert materials, deemed fairly safe to use in the medical field (surgical tubing and so forth). Maybe that was a big mistake, but most of the concerns I hear voiced right now amount to wild guesses. "Look at the increase in things like obesity, or cancer! Maybe it's the microplastics!" There are MANY other possible causes in play, so jumping to the conclusion microplastics are the next cigarettes is .... unreasonable.

> Also, it was pretty racist with the whole English thing.

The "whole English thing" is about having someone that it is possible to communicate with in English. I'm fairly sure they would be OK with speaking English through a translation system, as long as it didn't get in the way of the communication.

Also, as someone that is from a non-English-speaking country: Somebody that isn't able to operate in English is unlikely to be able to do good open source work. English is the language that is used to communicate about programming globally, and anybody that does not tap into that communication is at a very large disadvantage. The overall situation is unfortunate, and it can be argued that the situation is systemically racist - but it isn't individually racist to require the ability to operate in English. Fortunately, AI tools are continually making it easier to operate in languages you don't speak, so this disadvantage will become less and less of a problem.

One of the things people keep overlooking with skilled trades is the physical labor required. I know quite a few people with an I.T. career they're in largely because they have physical issues or limitations (bad back, for example) and it's not realistic for them to just hop over to a skilled trade like a plumber, where they're going to have to crawl under cabinets or into crawlspaces under porches all day long.

If you're still young, absolutely. It's not a bad time to consider a skilled trade to make a good living and escape the AI threat. But for the middle aged and older folks? Not always realistic.

This has always been the thing of science fiction (Star Trek universe being a great example). There's this idea that humanity will automate and invent so much technology that nobody has any work left to do.

In realty? Nobody on the planet has ever reached this point before in history, to prove it really happens.

All I *do* know is that in my own life experience, EVERY time I use a new tool or gadget that promises to save me time and make my life easier? Whatever time it saves is immediately claimed someplace else. In many instances, the claimed labor and time savings turns out to be artificial. (EG. I've owned several Roomba type robot vacuums. Even though each one I tried was an improvement over the old one? None have ever truly eliminated the time and labor I would have spent vacuuming by hand otherwise. What inevitably happens is they don't get certain areas clean enough, so I have to go back and deal with those myself. Then, they need maintenance and cleaning themselves, which requires a lot of time taking them apart and getting lint and hair unwound from places it wasn't supposed to be. Next, something wears out or breaks, costing me more money AND time to repair it or replace the worn part.)

I think all we might really see is some temporary pain around unemployment as we collectively re-think what kind of labor/work we still want to pay people to do.

AI clearly can't do ANYTHING physical, so that severely limits what jobs it can take from humans. Robotics can do physical tasks but we're still FAR from it handling random tasks that humans do easily. Our robots tend to be purpose-built. One may be able to scan inventory on store shelves, but it can't go clean up a spill in an aisle. Another may claim to do security duty but probably has to call in to humans if anything bad happens to it.

That's why it is using it to look over drafts. You take the AI output and then go through its suggested edits or concerns. Essentially it functions as a very fast, half-way decent proof-reader. For example, my spouse looks over a lot of drafts of things I write, but if I give her a 15 page paper, there's good work on the first 8 or 9 pages, probably a joke on page 10 about my notation, one or two comments on the next page, and then a few comments on the bibliography. Claude is not as good a proof-reader as she is, but Claude doesn't poop out after a few pages. And Claude is much faster. So reliability goes up without having to rely on the AI being reliable itself.

And sometimes it catches subtle things. For example, not too long ago, I gave a paper draft to Claude, and it noted in the bibliography that the journal volume number seemed unlikely given the year; it turned out that the volume was correct, but I had actually written the wrong journal down in the entry. Another time, Claude noted that one entry in a bibliography was from a "borderline predatory" (its words but it wasn't wrong) journal, and since I wasn't using it for much other than as part of a list of "hey, here's prior literature" and the paper in question didn't prove anything interesting, I discussed just dropping it from the citation list. Claude then without me asking, pulled up preprint I already had on the arXiv and noted it saw that paper cited there and suggested we remove the paper there also. I didn't do so, but this is the sort of thing that it can suggest. One still has the human make the decisions, but the input itself is useful. And for the most part, verification by a human for most issues is much faster than finding the issues one self.

Not really. It turns out that specialized models do not do as well as large general models. This is essentially a variant of the Bitter Lesson https://en.wikipedia.org/wiki/Bitter_lesson. In this case, there's other reason for it also. There's a lot of reasoning science and math discussions in non-specialized training data. For example, one paper I wrote with two coauthors started as a Reddit discussion. But there's also just a lot of subtle interaction that specialized models don't necessarily get. Math is easier to understand if one has a good command of physics. Engineering is better if one has math and physics. Etc. And sometimes a technique gets directly pulled over. For example, in the large OpenAI paper drop, one of the graph theory problems relied on a trick involving a weighing on the graph that involved complex numbers designed to cancel, where the trick came from a physics context that did something similar. A math specialized model would have been unlikely to find and import that trick.

Right, and the line here isn't clear. One take on NFTs is that they are nominally a tech but functioned purely as a commodity bubble. Or that NFT is a tech but it was used to make commodities which had a bubble. And the line between tech and commodity was blurry historically. When Spain entered the "platinum age," that was based in part on how to make nearly pure platinum was a tech, but the platinum itself was a commodity.

Since per a token output seems to be getting better also for open-weight models, this suggests this is more than just a business decision. And some of it is specific changes; we know that larger models with more training data for example will often produce higher quality output. But also aside from scaling, there have been architectural improvements. For example, one insight that occurred was that having lower precision weights which allowed more weights was often a positive tradeoff. That lead to changes in both the software and hardware ends, as the GPUs were modified as a result. Similar, there's been changes in what functions are used for the feedforward function.

The business case for LLMs is a bit broader than that. They don't need to replace that many workers if they make enough workers more efficient or more reliable (such as simply having LLMs look over drafts). There's also at this point, clear usage cases for scientific and mathematical research but that's using the very high end models, and admittedly is a pretty small market, not enough to justify massive expenditure at this level. The tulip example is interesting; what you are touching on their is the distinction between a "tech bubble" (like railroads or internet) and a "commodities bubble" (tulips, housing, Beanie Babies). Note that the degree to which the tulip bubble was real is questionable. See https://theconversation.com/tulip-mania-the-classic-story-of-a-dutch-financial-bubble-is-mostly-wrong-91413, and the line between a tech bubble and a commodity bubble itself can be blurry (see for example Bitcoin and NFTs). One of the difficulties here is that the LLMs keep improving, and the cost per token is going down, and the amount one can do per a token is going up. But that's happening in part due to massive investment in both infrastructure and research. How sustainable any of that is is unclear.

There are at least four problems with this. First, linear algebra is from a math standpoint insanely powerful. For example, all of representation theory https://en.wikipedia.org/wiki/Representation_theory is using linear algebra to understand groups. Much of the study of modular forms https://en.wikipedia.org/wiki/Modular_form involves understanding the behavior of modular forms over SL_2(Z) which has heavy linear algebra elements to it. People who see linear algebra as somehow a weak thing are often people who took a single semester of it and didn't do much else with it.

Second, the idea that these AI systems are just linear algebra isn't accurate anyhow. While there's a lot of linear involved, a major part of how transforms like LLMs work is the intervening non-linear layers, which use ReLu https://en.wikipedia.org/wiki/Rectified_linear_unit, Softplus, or a similar function. So, summarizing this as just linear algebra isn't accurate. It is linear algebra with some spice.

Third, even if this were somehow just linear algebra, that shouldn't therefore be a reason to think that there's any limits to what is going on. At the end of the day, the empirical success is what matter, and systems which can do things like prove the quasi-Riemann Hypothesis or solve other major math problems are pretty impressive regardless of what is under the hood. So, if one does think this is just linear algebra, that's not a reason to dismiss the systems but a reason to increase your estimate of what can be done with mere linear algebra.

Fourth, slowing down connections like this may make crashes *less likely.* In particular, the main way we could have a serious bubble here is over-investment. Delays including power grid delays force bottlenecks on the investment level, and make that sort of problem less likely.

We may be in a bubble here. Certainly some elements look very bubble-like. In the late 1990s, during the dot-com boom, one had people making perfectly normal businesses but just tacking on "on the internet" to it and it would get millions of dollars of funding automatically. See for example pets.com . And we're seeing some of that now; we've had a whole bunch of businesses see their stock price shootup just because they mention AI. So we may be in a bubble, and it may crash now, or might crash never or decades from now. (Arguably the US had an 80 year automobile bubble that didn't crash until the late 2000s which then forced the auto bailout.) But the other thing to keep in mind is that a crash does not mean the tech goes away. In 1873, the US had a devastating collapse of the railroad bubble, and yet railroads continued to grow, with the amount of track not peaking until around 1910 https://transportgeography.org/contents/chapter5/rail-transportation-pipelines/rail-track-mileage-united-states/. Similarly, although the internet bubble burst, by 2015, one had more internet use than even the most internet hyping person in 1998 would have likely dreamed of. A bubble bursting does not mean the tech itself goes away.

The set of problems that only a human can solve is getting rapidly smaller. The set of puzzles that a human can solve, that an AI cannot, and which take a small amount of time for a human and aren't highly inconvenient may already be empty.

Jumping a car with compressed air isn't a technological miracle. It is a useless stunt, and one which if went into production cars would be dangerous to both those in the car and people around them.

Geez people.... Regardless of who is President, the issue here is worth taking a good look at. $2 billion is a LOT of money to burn through each year! And this is a program that arguably went FAR beyond what people expected it to do (ensuring libraries and schools had an afford Internet connection). Handing out free hotspots to people? Yeah ... I'd rather not fund THAT with my tax money. Just use free public wifi like the rest of us do.