Technology Advances

[Jurassic]

> The question will be whether another monolithic and independent financial system coming from the east will get pumped with enough additional theta that it gets punted by a decade



@chiral3 ?



also why the negativity on "big data" - its growing all the time and hasnt been analysed yet?

[nikol]

Cheap chips lead to 'collect everything' -> BigData

BigData analysis require ANN (as they call it AI). As it is impossible to analyse data with 10k variables use of ANN (or similar ML techniques) is inevitable.

Let's face it - old things change their names so it becomes hype and sells:

Statistics ~ BigData

ANN ~ AI



China, Russia, Korea and Japan thing:

It is about change of trading patters - from sea routes into continental. Straits Hormuz, Singapore, Gibraltar, Suez-channel are not key points anymore. Look at global energy flows: Hormuz controls 80% of oil trade, Singapore 60%. When China-Russia penetrate whole continent with high speed container railway connections (+Northern Sea Route) all will change at once. These tests have commenced last few months and show significant reduction in cost and time of logistics between Europe and Asia (Africa is to come). Both Korea's despite resistance from US military base standing in between will agree on shared use of railway and port of Busan will get reduced load. Bridges Lazarev-Sakhalin, Sakhalin-Japan will connect Japan with railway for the first time ever. Imagine, train from Tokyo to London. Boom. New world. (I did not mention US purely accidentally. It stands behind its sanctioned walls)



By the way this redraw of the global map will demand a lot of technological changes in how railway transportation is planned, handled, optimized.



Upcoming market crash (the change I have pictured above will go through massive crisis) will lead capital into... crypto.

[EspressoLover]

> I still think there's potential in growth hormone secretagogues and things like this for anti-aging



The current market structure, regulatory regime and even broader culture is really ill-suited for anti-aging R&D. Metformin has pretty substantial evidence of slowing human aging. It's also dirt-cheap, super-safe, approved by every regulator, been used for decades, and has basically no side effects. If we can't even get the medical system to push metformin, why would anyone spend the resources and risk to develop some next-gen speculative compound.



> admittedly i read this fast but none of these possible breakthroughs seems to have happened



Over the past few decades most of the major hype has pretty much disappointed. There really hasn't been many breakthroughs, but a lot of things have improved so incrementally that they're barely noticeable. Computers, networks and consumer electronics are the headline. But even outside this there's been a lot of improvement. Cars are way safer and more reliable. Manufactured products are dirt cheap and made better. Cancer survival rates have steadily improved. Food is cheaper, fresher, more varied, available year-round and much less likely to make you sick. Air pollution and water pollution has gone down a lot in most developed countries. LED's have improved lighting along every dimension.



If I had to guess the near future probably looks a lot like the recent past. I doubt we'll live to see a period of breakthrough after breakthrough, like 1890-1960. But most things will probably just keep slowly and continuously improving. For example, I think fully autonomous cars are decades away. But we'll probably get widespread "super cruise control", and each year it will probably be able to do an increasingly higher percentage of driving. However getting from 95% autonomous to 100% will be a total bitch.



Quantum computers won't have much an impact anytime soon, but Moore's law will keep delivering way longer than anyone expects. Sell the hype, buy the quiet engineers in the backroom who are too busy and diligent to bother with PR.

[Jurassic]

I find it interesting how bill gates saw that the microprocessor would change everything and decided to write software for it.



@espressolover I really think that the automation of driving will be painfully slow and other technolgoies will appear before it, too many variables, too many consequences.



do you guys agree with peter thiel that innovation is slowing?

[billWalker]

Nikol, do you have some kind of source on rail transport through central Asia being cheaper than shipping?



I know that rail freight in the US is among the most efficient in the world (unlike passenger rail in the US), and according to Vaclav Smil, rail in the US is 5-10x more expensive than shipping.



I don't doubt the geopolitical and economic implications for Central Asia itself will be significant, but I hadn't seen any evidence that high speed rail would be competitive with giant container ships.

[nikol]

I am not in the market today, but 3 years ago delivery of wood pellets from region of Ussuri river to address in Europe was was more expensive by about 10-20% by rail. Since price formation of Russian state monopoly RZhD (=Russian Railway) is complete enigma, I would guess that to win the logistics flow they will do everything possible. Time saving is proven thing.



My personal working experience with Russian logistics is that initially they give you fair price, but when steady flow begins to move, they optimize further and give you much better price. It happens mostly due to great "unknown" (risk) at the beginning which is priced in. When this unknown becomes known all price chain drops.



Everything includes:

- cooperation with China and Korea (who are very interested in this venture)

- efficiency improvement is not the question (look impressive video of fully automated sea port Qingdao )



Visualize 2 paths of container delivery from Address-Korea to Address-Europe:



A. sea.

Address-Korea, on-load, transport, offload to ground, sea port, customs, ship load, shipment (~50 days), offload to ground, customs, on load transport, transport, offload, Address-Europe.



B.railroad.

Address-Korea, on-load, transport, offload to rail-platform, change of rail at border of Poland, load to transport, transport, offload, Address-Europe.



[I assume here that by end of the year N.Korea will provide the railway corridor for S.Korean goods]





Not yet what I envision, but you get the flavor:



FESCO reduces delivery time from Republic of Korea to Russia from 23 to 15 days



FESCO and Hyundai Glovis Jointly Launch Regular Cargo Transportation from Busan to Hyundai Plant in St. Petersburg





If you object and tell me about advantages of price formation in free market, I would fully agree. But why do we have sanctions, tariffs, tax subsidies and all this trade war today?



PS. FORBES: The Hidden Economic Rationale Of China-Europe Rail

[dws]

I have great experience and knowledge with your advance technology post from this forum section. keep sharing more tech advice.

[Jurassic]

>Cheap chips lead to 'collect everything' -> BigData

BigData analysis require ANN (as they call it AI). As it is impossible to analyse data with 10k variables use of ANN (or similar ML techniques) is inevitable.



@nikol not sure what you meant by the second sentence

[nikol]

example:



The data array with ~10 variables x 1 billion records is possible to analyse manually.



However, if you have 1k variables x 10 million records (same amount of data in bits), then manual analysis is very difficult. Neural Network techniques can help.



here is an example of such problem (Bosch at Kaggle):

https://www.kaggle.com/c/bosch-production-line-performance

[contol]

cool