Technical Idol Totalled, Stache Under Pressure!
19 posts • joined 28 Sep 2010
All right, bombastic bob, you have an increased risk of being asked to provide reason on this site.
You claimed two things:
1. "CO2 is released from water as it warms [...]" This is probably true, since at least this PDF https://sites.chem.colostate.edu/diverdi/all_courses/CRC%20reference%20data/solubility%20of%20carbon%20dioxide%20in%20water.pdf states that mole fraction of CO2 in liquid phase drops from 0.000814 @15°C to 0.000704 @20°C (both read from the last column at 100kPa, since one atmospheric pressure is 101.325 kPa). This may be different for fluids like seawater with salts and other gases and higher pressures underwater and maybe saturation has not been reached at all, but I am willing to accept your claim for now.
2. "CO2 only absorbs black-body IR radiation that corresponds to temperatures below -50F and above about 140F [...]" (-50F~=-46°C, 140F~=60°C) This is the part where you could provide reason to me: The following PDF https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/wea.2072 provides a red curve of CO2 absorption on the second page. It shows a peak @ 15µm. The black body emission spectral energy density for 0°C and 20°C on this graph http://www.giangrandi.ch/optics/blackbody/blackbody-ambient-large.png at 15,000nm (15µm) is not far from its maximum.
Did I make a mistake?
From Goodfellow, Bengio, Courville (2016) - Deep Learning:
on which common computer models learn: 10^4 - 10^9 (e.g. images). // Humans: when we are awake: full-time video, audio, tactile, smell, etc.
Connections per neuron
Computer model: 100 - 10^3 (there are outliers, but not commonly used) // Human: >10^4
Computer model: 10^6 - 10^7, also, neuron models are simplified a lot. // Human: 10^10
There is a lot we cannot efficiently model in silicon and more we don't know about individual neurons and their local interaction. One problem: we want to be able to transfer learning (topology, weights) between computers, so we have to use digital circuits.
Size is limited by computation speed, and well-parallelizable. I hope that Intel releases Stratix 10 MX soon, and not only to national interest buddies in the US. I think the work that can be done by current neural networks is impressive.
FPGA reconfiguration at runtime: It's being researched in academia, see scholar.google.com "partially reconfigurable FPGA"
Some state-of-the-art FPGA by the big vendors can already do this, support in their design suites is less-documented and incomplete, by my last information (1 year old).
I understood that these theories posit:
In the long run, as information becomes available, speculation will drive asset prices will toward asset real values. The players that do not (net)speculate in that direction will lose their ability to speculate for lack of funds.
I guess that is true. I think my father also told me that when he started speculating.
How little latency in speculation is needed to enable this driving of asset prices to their real values?
In the subprime bubble, some players suspected early that those houses were worth much less than consensus players assumed. And they were able to act on it, betting short months/years ahead, getting rich when the bubble imploded. These people did their homework, and acted at the right time for the long run to vindicate them.
I do not see how low-latency/high-frequency/high volume trading enables better and earlier homework, preventing large bubbles from occuring.
In fact, it enables those who do not do such homework to ride the bubbles. They can escape with less losses than those not doing such homework and not having access to low-latency trading (non-investment banks, fund managers ...). They can capitalize on the near-random fluctuations of the market and avoid the worst part of bursting bubbles.
I wanted to make an image of my newly bought Lenovo X220 tablet system, because I wanted to migrate to a separately bought SSD.
Lenovo recommends to use Acronis True Image (among others) in their handbook. I didn't want to take the time to make a working linux live flash drive, so I decided to throw my money at Acronis.
Of course it didn't work. (Same problems)
So now I'm deleting the old installation, made a linux stick and will install a free windows license I got from my employer, wishing I didn't have to install their shitty windows software.
I was going to buy the first serviceable Android tablet with
- a screen of 10" or more
- Google Marketplace access
- good stylus input. (E.g. a larger HTC Flyer)
I'm buying an Asus Eee Slate now, and hope for good Linux drivers and software in the future.
Does anyone here know any Linux tablet developments?
to pay for university. Until I fell from 36ft height onto a steel floor.
The climb down from the turbine should have been (relatively) secure since I followed regs and used protection equipment. But the anchorage in the wind turbine tower was installed incorrectly and bent open when my hands slipped on the oily ladder on a cold December morning.
In Germany you can't sue your employer for negligence, since there is a "Gesetzliche Unfallversicherung" that insures him against all your claims, unless he tried to injure/kill you on purpose. In return, the "Gesetzliche Unfallversicherung" is supposed to pay for an equivalent education if you can't do your job anymore and compensate you for permanent consequences of work accidents.
In my case, that means no help in getting an education (since I was still a student at the time of the accident) and 400€/month for not being able to walk without a crutch, two smashed elbows, some nerve damage to the arms and higher back pain when my back is upright. The courts still have to decide on that, though.
Nothing for about 20 month of hospitals.
At least I can type, and luckily my parents pay for food and room while I finish university.
I recommend against believing you will be cared for if anything happens to you while working in Germany. I don't know about other countries.
Biting the hand that feeds IT © 1998–2020