-
Posts
6,774 -
Joined
-
Days Won
3
Content Type
Profiles
Forums
Events
Everything posted by rkbabang
-
Bust up Your Paradigms or Get Left Behind
rkbabang replied to DooDiligence's topic in General Discussion
It appears that the revolution isn't AI at all. The big thing here is increasingly powerful alternatives to CPUs. CPU's have hit a dead-end and so the whole industry will evolve sideways. The GPU is only the beginning. We are going to see more and more very cheap alternatives to the CPU that outperform CPU's by orders of magnitude on specific tasks. The alternatives to CPU's have been going on for decades. I believe these chips are called ASIC's. Xilinx is one of the companies that make these. Altera is another one. One very good example of this is that in mining for Bitcoin, there were (are?) specialized computers that use ASICs and only do one thing...mine for bitcoins. They are incredibly efficient at this and are WAY faster than GPU's...Bitmain's "antminers" are a good example of this. You are mixing up terms a little bit (I'm an ASIC design engineer) Xilinx and Altera make PLDs (Programmable Logic Devices) which can be custom programed (even on the fly to a limited degree) to perform an application specific task/algorithm more quickly and power efficiently than a CPU could do by executing software. These are much faster than using a CPU to perform the same function, but not as fast as nor as efficient as a pure custom designed ASIC would be. ASIC is just an Application Specific Integrated Circuit. Any chip which isn't fully programmable and turing complete could be called an ASIC. The line is a little blurry sometimes as I've worked on plenty of chips which would be called ASICs yet had embedded CPUs inside them, yet they are programmed internally so that the chip itself only performs an application specific function and can't be reprogrammed from the outside by the customer (who may not even know there is a CPU in there). Basically when you are optimizing for programmability software run on a CPU makes sense, when you are optimizing for power efficiency and speed a custom designed circuit makes more sense. And of course the custom logic can interact with and control the CPU and vise versa A GPU has specialized circuitry to perform a math on matrices quickly, yet still be somewhat programmable which makes them useful in 3D graphics and simulating neural networks. I'm not sure AI will ever use ASICs rather than GPUs, completely, but a GPU with self programing PLD functionality would be interesting. The reason bitcoin miners moved from CPUs to GPUs to PLDs and now to fully custom ASICs is that bitcoin mining is just doing the exact same type of calculation over and over again with no need for programmability, so it is a perfect fit for a fully custom circuit. Turing complete CPUs can do anything the other chips can do, just more slowly and consuming more power. A fully custom ASIC does whatever it was designed to do more quickly and power efficiently. A PLD is like a large array of transistors which can be re-wired internally, it isn't as easy to reprogram them as it is to write new software for a CPU, but once configured it will do a specific function more quickly and efficiently than a CPU (but not as quickly or power efficiently as a full custom ASIC). Now if you think about it a CPU is just a type of ASIC, it is transistors on a chip wired in a specific way. So a CPU itself can be implemented as fully custom designs or programed onto PLDs. My senior project in college was designing a CPU with a custom instruction set and programing the design onto a bunch of Altera PLDs (it took 6 of Altera's largest PLDs at the time to fit my CPU design). It was nowhere near as fast as implementing the design in a custom ASIC like fashion, but it allowed a college student like me to implement my design in less than a year, build it, and start running programs on my processor in the lab. This is in fact how PLDs are often used (as prototyping devices), they are also used when the expected number of parts needed will be small. If you only need a few hundred parts and they don't need to run very fast you are probably better off using a PLD, but if you are going to sell a few million chips, or they need to be as fast as possible you are probably better off designing an ASIC. A CPU or a GPU are really just special kinds of ASICs which can run software. My apologies if this was too rambling and confusing. -
Fixating on metrics is at least an investment thesis. The worst is fixating on your purchase price, which is irrelevant for anybody but yourself. You & LC are super right here! I have a really hard time paying more than my last execution price for an issue. Kind of crazy, "I'm pretty confident that it'll be worth a lot more in a decade but I don't want to pay a premium above the benchmark I've established in my head." Or even worse is that you don't buy it at all, because after you have convinced yourself that you want to own it, you then want to wait until it is back down to the price it was when you first added it to your watchlist. You just can't bring yourself to pay that premium, as if the universe owes you the opportunity to buy it where it was a few months ago when you first started looking at it. Ughh. The human brain sucks sometimes. These are the default modes of thinking unless you consciously override them which takes some effort.
-
Two interesting stories about Ethereum. Singapore will be issuing its currency as tokens on the Ethereum blockchain. http://www.trustnodes.com/2017/06/07/singaporean-dollar-tokenized-ethereums-blockchain-monetary-authority-singapore And Putin had a meeting with Vilalik Buterin https://www.bloomberg.com/news/articles/2017-06-06/putin-eyes-bitcoin-rival-to-spur-economic-growth-beyond-oil-gas
-
+1. Great post. I've done this so many times I can't even count. I convince myself that something is cheap. I buy it, and after another few days or a week of research decide that I was fooling myself and I sell it again around where I bought it for (sometimes for a small gain or loss) then move on. Sometimes things are cheap for a reason and the market has already figured it out even if you haven't. I think efficient market theory is BS, but at the same time you have to admit that the market is often correct.
-
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
Interesting in that some states that use a lot of coal, like HI, KY, CO, NM, Hybrids pollute the least of all, whereas most states full electric is the least. Some states which use very little fossil fuels other than Natural Gas such as CA or NH the difference between gas and EV is extreme. I can't find any state where gas wasn't the worst (I didn't look at them all though). EDIT: It looks like the two extremes are Vermont and West Virginia. -
Bust up Your Paradigms or Get Left Behind
rkbabang replied to DooDiligence's topic in General Discussion
Nice article even though I only understand a tiny fraction of what they're talking about. "The AI revolution has arrived despite the fact Moore’s law – the combined effect of Dennard scaling and CPU architecture advance – began slowing nearly a decade ago. Dennard scaling, whereby reducing transistor size and voltage allowed designers to increase transistor density and speed while maintaining power density, is now limited by device physics." Did the author mean not instead of now? No he meant now. Device physics is now a serious limitation to CPU scaling. I wonder the same thing myself. I missed NVIDIA too. I don't know, Apple keeps what it is working on secret, as I'm sure you know, but there are rumors. Apple Developing 'Apple Neural Engine' Chip to Power AI in iOS Devices I get some pretty good articles emailed to me from Linkedin. Maybe because I follow a bunch of tech related companies and people on LinkedIn. I read something pretty interesting about what Adam's could have meant by "42". 42 is the ASCII code for the asterisk symbol (*) on the keyboard. Which is sometimes used as a wildcard symbol when searching or in some computer code such as regular expressions. So what Deep Thought was actually saying is that the meaning of life, the universe, and everything is whatever you want it to be. -
I don't see how your views are contrarian at all. The value of these things is completely speculative at this point. I've never seen anyone dispute that. Even the most ardent supporters and evangelists are talking about what blockchain technology could become not what it is already. Right now these things are used very little by very few, other than for speculation purposes. As JB85 said above it is an asymmetric bet. That is why I took a small amount of money and purchased a basket of them. That small amount has already become a measurable position in my portfolio, but I also understand that this is a speculative bubble and it will likely crash back down soon. ETH is still flying high, but BTC is already down significantly off its recent high.
-
Bust up Your Paradigms or Get Left Behind
rkbabang replied to DooDiligence's topic in General Discussion
Excellent series. Thanks. This is a good article as well: "Software is eating the world, as Marc Andreessen said, but AI is eating software." https://www.linkedin.com/pulse/ai-eating-software-jensen-huang -
So if you think this will represent 2% there is still a lot of growth left. If you think it will be 15%, or 51%, or 80% there is a hell of lot of growth left. :) Exactly. If you think there is any value at all, either as currency or store of value, then even at present prices, crypto in general is one of the most asymmetric bets of all time. I can understand people saying bitcoin will go to zero. What I find odd is when people say the fair value is 50b to 100b. Your basically saying it should be exactly 1% of world money supply which seems oddly specific +1. I agree. It is either wildly undervalued or not worth anything. 80% would shock me, but I expect that 10-20%+ of all the world's economic transactions will take place on blockchains within the next 15-20 years. Although I suspect my prediction will prove to be grossly inaccurate in one direction or the other.
-
So if you think this will represent 2% there is still a lot of growth left. If you think it will be 15%, or 51%, or 80% there is a hell of lot of growth left. :)
-
Since I wrote this 21 days ago, Bitcoin now has a $45B marketcap and Ether $19B. There are 7 with caps >$1B, 31 above $100M, and 250 above $1M. The total space is over $90B.
-
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
Yes I realize not everyone wants to drive the same car for 15 years and NH is one of the cheapest states to own a car , I paid more than double for insurance when I used to live in MA, so maybe $2K per person is a more reasonable threshold. -
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
I wanted to put some real numbers on this, so I thought I'd figure out what my current two cars cost me per year in total assuming I keep them both for 15 years. I purchased a 2007 Hyundai Elantra in 2007 and a 2006 Toyota Sequoia in 2011. The rest are my best estimates trying to overestimate not underestimate and then round up. 2007 Elantra estimate from 2007-2022: Purchase price: $15,000 3 sets Tires: $3,000 2 timing belts / major services: $4000 Estimate Other Problems (brakes/struts/exhaust/etc): $8000 Gas @ $35/week: $27,300 Oil changes 2/year: $1500 Inspection stickers: $525 NH Registration: $3000 Full coverage Insurance: $9,000 ----------------------- Total for 15 years: $71,325 call it $75-80K 2006 Sequoia estimate from years 2011 - 2025 (I don't put anywhere near the number of miles on this per year as I do the Elantra, but it is more expensive to fix/fuel/insure/etc, so these numbers take that into account) Purchase Price: $25,000 2 sets Tires: $3,000 1 timing belt / major services: $4000 Estimate Other Problems (brakes/struts/exhaust/etc): $8000 Gas @ $40/week: $31,200 Oil changes 2/year: $1500 Inspection stickers: $525 NH Registration: $4000 Full coverage Insurance: $12,000 ----------------------- Total for 15 years: $89,225 call it $90-95K So $165-$175K for 15 years which is $11-12K per year. Damn, cars are expensive even if you buy them infrequently and drive them a long time. So to save me 80% the service for my wife and I would have to cost us about $2200 per year. -
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
If you live and work in NYC it is certainly possible to live without a car today, many do, but that isn't the case for the vast majority of Americans. Cars are necessary and having your own vehicle always available at a moments notice is very convenient. I wouldn't give up my cars for a 10-15% savings, it would have to be an 80+% savings before I'd even consider it. There is no way to achieve that if human drivers need to be paid. Also, I don't want to ride with strangers in the car, so even Uber as it currently exists is unacceptable for that reason alone even if it was cheaper. -
I disagree, with your second point, not your first. Yes a blockchain needs a token to give incentive to secure it, but I don't see bitcoin as the clear winner. I'm leaning more toward Ethereum now, and not just because it has grown to be my largest crypto holding (by a factor of almost 2 now), but because it has a road-map which includes privacy changes and it is more functional being turing complete. The Bitcoin community can't even get together to fix the blocksize issues and other scalability problems. Right now I think BTC is surging because it is the gateway currency, i.e. the easiest way to acquire the others is to get BTC and then exchange it on an exchange or with shapeshift, etc. Coinbase is the only way that I'm aware of to acquire ETH or LTC with US dollars directly and I am not aware of a way to buy any of the other altcoins except with BTC. I have a feeling that will change rather quickly. You will see ETH ATMs and other ways as ETH becomes more popular. The Rise of Ethereum (http://www.nasdaq.com/article/the-rise-of-ethereum-cm794082): "Over the years, Ethereum has garnered admiration and evoked immense interest across enterprises. Accenture observes, “Every self-respecting innovation lab is running and experimenting with Ethereum, including IBM, Microsoft, JP Morgan, and the R3 consortium.” Ethereum is supported by Microsoft Azure and Alibaba Cloud to encourage innovation and adoption of the technology. However, the efforts remained scattered until the formation of the Ethereum Enterprise Alliance (EEA) in 2017. The EEA brings together enterprises, startups, academics, technology vendors, and experts to work on Ethereum as an enterprise-grade technology. Its founding members include names like Microsoft, Credit Suisse, Intel, JP Morgan, UBS, Santander, ING, CME Group, BNY Mellon, Accenture, and Wipro. Some of these names (such as Microsoft, JP Morgan, and Santander) have been actively involved in exploring the blockchain technology in the recent years, while others are looking at becoming a part of the revolution that’s underway. Microsoft believes that Ethereum is “evolving to address the needs of enterprises globally. Focusing on requirements like privacy, permissions and a pluggable architecture while retaining its public roots, Ethereum continues to widen the scope of what developers, businesses and consortiums can achieve.” During Consensus 2017, Ethereum Enterprise Alliance added another 86 members—taking the total count to 116. EEA now represents well-known names from across different regions and industries, such as Mitsubishi UFJ, DTCC, Deloitte, Samsung SDS, Infosys, Toyota Research Institute, National Bank of Canada and Merck KGaA, among others."
-
There would still be a mining reward, which would come out of the transaction fees alone. Also the date is 2140 not 2040, the likelihood of bitcoin still existing in exactly its current form without being hard forked into something different at least once along the way, and it still being the dominant cryptocurrency, 123 years from now seems slim to me.
-
I was just thinking about how this would work in an unregulated market? These companies would design their apps to use Ether directly inside their apps, but raise money by issuing tokens on the blockchain which represented an equity share of the company. These tokens would act just like stock does, but traded as easily as a cryptocurrency. They could be traded on the blockchain exchanges, they could be traded directly from person to person outside of an exchange, the company could issue "dividends" in Ether to token holders, the company could buy back tokens, or do secondary offerings to raise more capital. Even non-tech companies could "go public" by issuing equity tokens on the Ethereum blockchain fairly easily. These tokens would have an intrinsic value in the same way that stock does today, because they would represent shares in the company. Of course since anyone could participate in these ICO events anonymously from anywhere in the world, this would be a legal nightmare in our current regulatory environment.
-
I found these articles informing. One question I had was why do these apps lately all seem to be issuing their own tokens on top of the Ethereum blockchain rather than using Ether directly in their apps. It is a way to raise money around US securities law. That is why they issue a token rather than a share of equity. From: The difference between App Coins and Protocol Tokens "Unfortunately due to securities regulations it is difficult for dApp developers to raise money on the blockchain through a public sale of unregistered securities. So while equity app coins might provide the most logical incentives for both dApp developers and investors alike, it is currently too risky to structure app coins in this way. That being said, public crowd sales are an irresistible source of funding for dApp developers and with traditional venture capital firms just starting to dip their feet into the world of digital assets, there aren’t many reliable alternatives." ------ And from: A beginner’s guide to Ethereum tokens width=606https://cdn-images-1.medium.com/max/1200/1*fG1euKN8v0-Q32gWGPuihA.png[/img]
-
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
I wouldn't ride in the vehicle if there were other passengers. Your service would have to be a lot cheaper than the car services. Autonomous busses don't seem very attractive to me, I would never give up my car for that. -
Fidelity to allow clients to see digital currencies on website
-
Zcash has been on my list to look further into, but so far I haven't. There are so many interesting projects out there that there isn't enough time to look into them all. Bytecoin is another one and I just heard about EOS recently which looks interesting too https://steemit.com/@eosio I really like the article Liberty posted above: https://stratechery.com/2017/tulips-myths-and-cryptocurrencies/ I think that if this is a bubble it is more analogous to the late 90s internet bubble. A lot of development, a lot of money poured into a lot of different ideas, but most of them will fail. It is probably too early to know which ones will be the Amazon.com's or the Ebay's and which ones are the pets.com or the Netscape's. I only own Ethereum(ETH), bitcoin(BTC), Monero(XMR), & Dash(DASH). And that is the order from my largest holding to my smallest at today's prices.
-
If you use coinbase it works like Gemini you have an account and need to trust coinbase. You could however use coinbase (or Gemini) to acquire your cryptocurrency and then send it to a wallet which you control the private keys, such as Jaxx (or another I've used is Bread Wallet, but I like Jaxx better now from a usability standpoint and the fact that it supports other currencies besides bitcoin and has Shapeshift integration. Bread Wallet is bitcoin only). coinbase and Gemini will report your transactions to the IRS, so if you use this method you should keep track of all of your buys and sells for capital gains purposes (you are legally obligated to do this in the US even if you acquire your bitcoins using cash at an ATM or from a private party). Here's a guide I found, but have not read through it, so YMMV. http://www.coindesk.com/information/ I did read "Mastering Bitcoin" by Andreas Antonopoulos which I'd recommend if you want a deeper understanding of how it all works. Not necessary to use it though. My best advice would be to open an account with coinbase or Gemini and make some small transactions to see how it works. Buy some bitcoin and ether, transfer it to your Jaxx mobile wallet, use shapeshift to convert some to Dash or something else. Get a feel for it. And yes, it might be just a bubble and you can lose money. Thank you very much. If I download it into a wallet and stored it on my PC do I need to have a fear of someone hacking my computer and stealing the wallet? If my hard drive crashed am I just screwed out of the bitcoins? Using Jaxx you don't need to worry about losing your wallet if your harddrive fails or your phone dies if you write down your backup phrase. It is a 12 word code that will restore your wallet to any device. So you can have the same wallet on your PC and your phone. Create it in one place and use the 12 world phrase to load it onto the other device. Any transactions you make with one device will be seen by the other because it is on the blockchain. In fact you can download Jaxx now at https://jaxx.io/ and create your wallet (it will have a zero balance for all coins) and see your backup phrase to see what I am talking about. Write that down and put it someplace safe, like a safe, a lock box, or even a safety deposit box. If someone finds it they can get full access to your wallet. As far as if someone hacks your computer or phone, you can create a pin code to open Jaxx so they would at least have to crack that, but yes some people keep only small amounts on their phone or computer and keep larger amounts on paper wallets or hardware wallets such as TREZOR or Nano Ledger. The great thing about these currencies is that using tiny amounts is exactly the same as large amounts. You can buy $10 worth of bitcoin in coinbase, send $5 to Jaxx, use shapshift to convert $1 to ETH and $1 to DASH. Create a paper wallet and put $0.50 in it, use a block explorer to verify that your paper wallet has the value, then take $0.25 cents back out. In this way you can experiment with all these things, get comfortable with them, and not risk any real money.
-
If you use coinbase it works like Gemini you have an account and need to trust coinbase. You could however use coinbase (or Gemini) to acquire your cryptocurrency and then send it to a wallet which you control the private keys, such as Jaxx (or another I've used is Bread Wallet, but I like Jaxx better now from a usability standpoint and the fact that it supports other currencies besides bitcoin and has Shapeshift integration. Bread Wallet is bitcoin only). coinbase and Gemini will report your transactions to the IRS, so if you use this method you should keep track of all of your buys and sells for capital gains purposes (you are legally obligated to do this in the US even if you acquire your bitcoins using cash at an ATM or from a private party). Here's a guide I found, but have not read through it, so YMMV. http://www.coindesk.com/information/ I did read "Mastering Bitcoin" by Andreas Antonopoulos which I'd recommend if you want a deeper understanding of how it all works. Not necessary to use it though. My best advice would be to open an account with coinbase or Gemini and make some small transactions to see how it works. Buy some bitcoin and ether, transfer it to your Jaxx mobile wallet, use shapeshift to convert some to Dash or something else. Get a feel for it. And yes, it might be just a bubble and you can lose money.
-
Rethinking transportation 2020-2030, Tony Seba, Stanford
rkbabang replied to indirect's topic in General Discussion
There are still horses and horse people who love them. Not many though, but almost nothing ever goes away entirely. -
If you want dead-simple. Open up an account with coinbase. It is a US based company and follows all US laws, reports to the IRS, etc. You can open an account, link your bank account and buy bitcoin, ether, or litecoin. You can use their mobile app as a mobile wallet to do all of this and to send/receive any of the currencies. If you want to stay more private I'd recommend downloading the Jaxx mobile wallet (or chrome extension on your PCs chrome browser). You can then obtain bitcoin, eth, or a whole host of others, and have it transferred directly into your wallet. The problem is how? If you have a bitcoin ATM near you (there are a few near me in NH) then you can put cash in and have it deposit bitcoin into your Jaxx wallet. Or you could find someone locally willing to sell you some for cash (sometimes these people advertise on craigslist, but be careful this sounds risky to me). Once you have bitcoin one way or the other you can use Shapeshift which is built right into the Jaxx wallet to convert bitcoin to other cryptocurrencies. EDIT: Here is a site which lists locations of bitcoin ATMs. I'm not sure how accurate or complete it is, but a few of the ATMs I know about are listed there correctly: https://coinatmradar.com/country/226/bitcoin-atm-united-states/ There are other wallets and exchanges, but those are the two methods I'd recommend.