nsx5200

It's interesting to see that the people that started "modern" AI are mostly not in the top right corner (AI maximalists). At least not on it's current LLM path (Lecun).

I remember when Satellite phones like Globalstar and Iridium came out, and people just extrapolate the growth/coverage to imply that cellular phone was going to go the way of the Dodos. Instead of going away, cellular phone coverage just kept building and increasing capacity. Satellite phones remained a niche product. Even now, the famed Starlink is still a niche product, when compared to traditional terrestrial products, despite all its success in displacing the incumbent in that niche. The issue with listening to business executives, even CIOs, that have little to none technical background in how AI really works is that they repeat the same mistake of taking the current capability growth, and project it out to infinity. In the case of AI, that is the end goal of super intelligent GAI. They will even adopt the mainstream "tech" buzzwords like "first principal" without actually apply the concept. Be wary of AI maximalists/minimalists that hand-wave a lot of the details and throws around a lot of buzz words. Any good engineer knows that the devil is in the details, especially in tech.

The most important part of that is this: "Joe Atkinson, PwC’s global head of AI [...]" Buffet said something about Barbers and haircuts.

Those are excellent counter-points. The real long-term issue for them now is the maintenance cost associated with these projects. In an effort to reduce cost & speed up development, proper feasibility study is not done, causing these projects to be non-self sustaining. A CCP version of the bridge to nowhere. All systems will have some slop, and I guess in their system, the slop shows up in this form. Hyper competitive has the second-order effect of involution. So it's one more issue that the CCP have to contend with. The primary incentive system is based on the mechanism described, so the corruption becomes embedded in their system. I do, however, like how the Politburo acts like a corporate BOD, and brings in experts to educate them on different matters. It is definitely a different system than the "west", with its own pros and cons.

On the topic of potential AI bubble: Ronny Chieng Investigates the Promises of AI, the Most Expensive Circle Jerk Ever

Wasn't sure if this belonged in the Politics or this thread. It's a bit long, so I highlighted the relevant mechanism and tried not to touch the politics as much as possible. I came across this video w/ Victor Shih that explains a lot of the mechanisms that drive the decisions that steers the Chinese economy. The part that is most relevant to China macro, IMHO, is the flawed incentive system that encourages the government to overinvest, sometimes absurdly, in various projects (real estate, rail building, etc). At around 1:04, Victor describes that mechanism: NotebookLM snippet "Here is a breakdown of the incentives: 1. Rent-Seeking and Kickbacks: The most realistic mechanism linking investment and promotion is the corruption and kickbacks associated with large-scale projects. ◦ When a large infrastructure project is initiated (e.g., building light rail), a significant amount of money changes hands through contracts for materials like cement and various contractors. ◦ A local official or governor can receive a large cash payout (kickback) from the contractors. ◦ This illicit money can then be used to pay off superiors, which "increase[s] your chance of getting promoted". This is described as a rent-seeking mechanism "

IMHO, I don't think Trump really care if he puts people at risk for his own personal gain. Just look at his actions during J6, even when the people around him urged him to take action. There seem to be enough China hawks in the administration and legislation that act as some counter balance to that tendency. Based on the actions to beef up forces in the Pacific, it seems like a fair assessment. It's possible that China/CCP can woo Trump enough to give up Taiwan during his upcoming visit to China. If China/CCP want to achieve their objective, they should break out Olympic-like shows for Trump. But it would make a stark contrast against the economic suffering that the Chinese people are currently undergoing, thus increase the chances for a coup/uprising in the future. Those pesky second-order effects...

Most of the "money" is "made" by the early adopters, that's why all the new grifters push their own crypto instead of Bitcoin. Thanks for your lengthy input arguing for Bitcoin. It helps to calibrate the narrative/viewpoint/myths/facts surrounding Bitcoin vs other cryptos. If you don't mind me asking, for my own calibration sake, you don't have a technical background in weeds of computing, do you (ex. computer programming, protocol design, algorithm design, computer/hardware engineering, etc. or even something in the hard sciences(specifically Physics))?

In regards to Japan and Taiwan, it doesn't seem to be that simple, as Trump just signed Taiwan Assurance Implementation Act to counter China. On the surface, based on the Whitehouse release, it sounds like a routine bill without real significance. But when you look at the text from the bill's sponsors, it's essentially an anti-China(CCP) bill. This administration is trying to shift the focus away from Europe towards China, as Trump, thinking the US as a continental power instead of a maritime power, try to contain China. If anybody follows Sarah Paine, she thinks the actions from this US administration is shifting the geopolitical center from the US back to Europe again as Europe strengthens their institutions in order to deal with the ongoing & upcoming global issues. Authoritarian governments like China, Russia, North Korea, Iran, etc., over the long time, have shown to be failed models. The more authoritarian, the slower the growth, if at all. We have seen how quickly China's growth slowed when the party becomes more authoritarian. We should invert, and avoid those authoritarian tendencies.

Thanks. That actually makes a decent amount of sense in that Bitcoin has niche use cases. Although... using Linux as an analogy undersells Linux. By a lot. Linux in the server/development space is actually displacing windows in that space. Even Windows adopted Linux starting in 2016 as WSL. The reason being that Windows/MacOS has a lot more focus and refinement in the UI space, whereas *nix tend to focus on safety and reliability, which is more important in the server space. Interestingly enough, Linux is also the core of Android, as well as unspeakable number of embedded devices. Combined, Linux deployment numbers easily dwarfs any comparison to Windows, MacOS, or even niche other embedded OSes. But that is a flaw in the analogy and not a flaw in the argument. Is there a particular space where Bitcoin, or even crypto-derived technology is displacing some incumbent fiat currency space, or is the direct competitor mainly to gold? If Bitcoin is that niche, then IMHO, there are some serious flaws to the premises of future widespread Bitcoin adoption. I appreciate your pushback/feedback.

So if the killer app, with all its advantages, is to move money around, shouldn't we see a hockey stick adoption in terms of aggregate dollar amount transacted per day/week/month? https://bitinfocharts.com/comparison/transactionvalue-btc.html#3y [wrong link. thanks, gfp] https://bitinfocharts.com/comparison/sentinusd-btc.html#3y Maybe there's a better metric that can be used to measure that, but I thought aggregate dollar amount seems to be the most reasonable to me out of the slew of possible metrics available from that list. It's possible that the source data is flawed. If so, what good source of aggregate data should I be looking at instead?

Just curious. What's are the "moats"/secret sauce for Bitcoin vs other cryptos besides first mover advantage? I guess the follow-on question would be, how would those moats erode?

Ran across this video that explains non-normal distributions that is prevalent in the world. Here's a NotebookLM summary snippet: "While many natural phenomena, such as human height, cluster around an average and follow a normal distribution, a significant number of systems in nature, technology, and society are governed by power laws. These systems are characterized by the absence of a meaningful average, a lack of inherent scale, and a much higher probability of extreme events than a normal distribution would predict." I believe it's essentially the same concept covered by Benoit (fractal guy)'s book: https://www.goodreads.com/book/show/665134.The_Mis_Behavior_of_Markets and the underlying model that powers Black Swan events. It also explains those outsized contribution by a few winners as seen in the investments made by Sequoia.

When TPU came out, it was ahead of its time. Depreciation is an accounting thing. We know that things can be fully depreciated but still useful in generating revenue. A lot of that depends on whether the semi fabs can continue Moore's law. If somebody figures out how to make compute hardware cheaper quicker, than those depreciation will speed up. Similarly, if they find out they're hitting some barrier that slows down the improvement of compute hardware, than those depreciation will slow down. Those barriers may include access to critical minerals, destruction of existing fabs due to war, etc... So the TLDR summary is that nobody really know, and 4-6 year guesstimate based on Moore's law and historical empirical data is probably a decent one, but still a guesstimate. Whether these companies are fudging these grey area numbers to 'cook the book' is a separate discussion. Just curious, does changing the depreciation schedule from 4-6 year really change your investment thesis that much?

Care to share your insight? TIA

Google seems to output an apt description, with my TLDR highlight: ""Economies of scale" refers to the cost advantages a single company gains from increasing its production volume, while "scale economies shared" is a specific business strategy where a company passes those savings on to customers through lower prices. The former is a general economic principle, whereas the latter is a deliberate tactic to build market share. Economies of scale Definition: A firm's average cost per unit decreases as its output increases. Mechanism: This is achieved by spreading fixed costs (like rent or machinery) over more units, but also through bulk purchasing, specialization, and greater efficiency as the scale of operations grows. Benefit: The primary beneficiary is the individual firm itself, which becomes more profitable and competitive. Scale economies shared (SES) Definition: A business model where a company strategically uses its economies of scale to lower prices for its customers. Mechanism: After achieving cost efficiencies, instead of just increasing its own profit margin, the company passes the savings to consumers in the form of lower prices or better service. Benefit: This creates a virtuous cycle: lower prices attract more customers, which increases sales volume and scale, leading to further cost reductions and deeper competitive advantage." These AI tools are like your personal tutor that never gets impatient, or a tutor that never finds your questions too dumb or too off-tangent to answer. The responses are near instantaneous to speed up the QA learning loop. IMHO, it's worth spending some time to learn to use these tools to the limits of these tools. Because in the world, you'll be competing against others that will have access to this tool. To compete without it is like the guy that brings a knife to a gun fight.

Current 'state of the art' AI is still some form of feed forward network. The training is fed back to update the weights, but in actual inference/implementation, it is expressed feed forward. The organic brain is more organic, so it feeds forward, back, sideways, etc. The TLDR is that structurally, organic brain is different than current computing AI. There have been some success with current compute AI, but it is still no squishy brain, so IMHO, there will be fundamentally different problems that these different types of 'brains' can solve. Even within the comparatively confined game like Go, which a statistic machine that has optimized moves up to the nth move in the future, Lee was able to gain some insight to beat AlphaGo once. Compute brains are still fundamentally statistic optimizing machines at the core, with some randomness added to it, whereas organic brains are not, for better or for worse. I think the trick is to recognize/learn the type of problems that is better suited for the type of brain to better solve the problem. Right now, people think compute brain will equal organic brain, and IMHO, I just don't see that happening without a breakthrough in compute brain's structure. If I'm right, then it'll take people some time to realize that.

Listening to the podcast series "What Trump Can Teach Us About Con Law", it's pretty clear that the system of checks and balance is set up to be adversarial to work. Each of the three branches are suppose to fight for their own turf and not cede too much of their power, with the House of Representative having a slight edge within the bicameral since they're the ones that are closest in representation to the people. Currently, we've seen that there are serious collusions between the legislative branch and the executive branch, and it seems like the Judicial is becoming more political than before as well. This does not bode well for the people. That's why I'm starting to gain a new respect for MTG and her willingness to fight the system, no matter which 'side' is in power. I may not agree with her reasons, but I can appreciate her fighting for what she believe is right for the people. We need more people that's willing to fight for the people, even if the system is against them. I don't want to name any names, but we are really all on the same 'side' with shared outcomes, so I urge people to debate about the merits of issues, and not which 'side' is winning or which 'side' is worse. There are good ideas on both 'sides', and there are not-so-good ideas on both 'sides'. Ideally, we pick the best ideas from both 'sides'. For @John Hjorth: Hover over the quote, and select that to cut/copy/paste it, even to a different quote.

A podcast series for those that find @Dalal.Holdings's wail on corporate misdeeds 'educational', especially when its fruits from the trees of HBS, McKinseys, etc: https://podcasts.apple.com/us/podcast/corporate-gossip/id1650850378 Corporate Gossip: The business podcast for gossip lovers Corporate Gossip explores the most absurd corporate scandals, the juiciest wall-street tea, and the not-so hidden secrets of famous business people. You don’t need to know anything about business to enjoy the Corporate Gossip Podcast, but you might learn something along the way!

Or just be a Booglehead and DCA through the highest and lows. Some people think it's going to run out of steam, but there seems to enough people out there gambling with cryptos, meme stocks that IMHO, it can still work for the next 10 years.

I bought some OTM LEAP puts for Google, as well as trimming the underlying asset as well. Going to keep a significant portion as a lottery.

Found on Reddit+Facebook, for those that still remember The Karate Kid from the '80's.

I did it for my wife's 401k about a year and a half ago. It might still be on-going, as I see this in her account: The transfer was through a third party, and that third party took its time. That caused the transfer to missed the deadline, so she only got 1% match. I don't remember exactly how long, but my poor recollection is that it took ~2+ months to do, so try to factor that in w/ the deadline. Like what you stated, Robinhood is a PFOF(that is how they make the big bulk of their money, thus the huge marketing effort to suck in money and induce unnecessary trades). I do feel like the execution price is not as good as Fidelity, so consider it a relatively a high frictional cost platform, and mitigate it by buying-and-holding on it. Their interface is very modern and very responsive so I do a lot of my pricing research on it before actually placing the actual order on other platforms. Other tidbits: The 3% bonus requires a 5 year hold, as seen in the screenshot(in the fine print). They don't pay interest on 'cash' in the retirement account, AFAIK, so you're forced to DIY w/ short-term T-bill ETFs like SGOV. The gold subscription does come with free $1k margin in the non-retirement account, so you can use that free margin to buy T-bill ETF to defray the cost of the subscription without incurring too much risk. They do offer limited margin and dividend reinvestment as well, if those are important to you. I would recommend reading other people's experiences from Boglehead and Reddit as well. https://www.bogleheads.org/forum/viewtopic.php?t=410277 If you have specific questions or features, I'd be glad to help you research.

New MEME stock spotted: https://finance.yahoo.com/news/forget-gamestop-meme-stock-traders-153809146.html Anybody playing this?

The CCP and it's internal politics are as opaque as they come, with announced and unannounced actions that can be interpreted both positively and negatively for Xi. That's why I'm hesitant to go down a discussion trying to evaluate the odds of such a change, but would rather focus on the economic ramifications of such change, and leave the evaluation of odds for a political change to individual investors. At the risk of going down the wrong rabbit hole, this is all I'll comment: Xi's position may look stable from outside, but when their de facto top military command leadership changes, many of whom were promoted by Xi. It's most likely there are major political moves involved. From ramping up on CCP politics, 'corruption' charges are just bogus charges to rid of people, and not actually use to tamper down corruption, as the whole system is built on bribes funneling to the top. This exercise has many similarities to reading tea leave, and I hate to propagate more speculations so I'm going to leave it at that.