Science and Technology links (January 5th, 2018)

  1. Regarding solar energy, Tyler Cowen writes:

    There is now a doctrine of what I call “solar triumphalism”: the price of panels has been falling exponentially, the technology makes good practical sense, and only a few further nudges are needed for solar to become a major energy source. Unfortunately, this view seems to be wrong.

  2. Laura Deming is a venture capitalist who has a “longevity FAQ” where she summarizes what is known about human longevity.
  3. Mitch Daniels is president of Purdue University and a former governor of Indiana. He writes that Avoiding GMOs [genetically modified organisms] isn’t just anti-science: It’s immoral.
  4. For web developers, having a Computer Science degree does not improve income, or if it does, the effect is small (about $1000).
  5. Year 2017 was the best ever for airline safety.
  6. Lixisenatide, a drug developed to treat type 2 diabetes, shows neuroprotective effects in a mouse model of Alzheimer’s disease.
  7. Only 6% of Britain’s territory is built upon, and only 0.1% is densely built upon. Yet people estimate much greater fractions:

    The story of Britain’s treasured green landscapes being gobbled up by greedy industrialists and developers is part of national folklore. The myth that we are “concreting over” our countryside plays to these anxieties, magnified in recent years by the social changes brought about by globalisation and new technology.

  8. The number 277232917 – 1 is prime. It takes about 30 hours to verify that it is a prime number. Prime numbers that are just off from a power of two by one are called “Mersenne primes”, and this latest prime number is the largest Mersenne prime number known to us. There are now 50 Mersenne primes. They are useful in cryptography though I am not sure this latest prime number will serve a useful purpose.
  9. A young postdoc has discovered a security flaw affecting all recent Intel x64 processors called Meltdown. The issue arises because Intel processors speculatively try to read arbitrary memory, and only later cancel the read when it finds out that the read is not allowed. It allows a program to bring any memory to the CPU cache where it can be read using some clever tricks. Given that processors cannot be changed and that most software won’t be recompiled anytime soon, this means that vendors like Apple and Microsoft have to issue fixes that will potentially slow down our computers. It seems like AMD processors are not affected. A related flaw (Spectre) is believed to theoretically affect most modern processors, including those in your mobile devices. Unlike Meltdown, Spectre is harder to exploit, but it is still being taken seriously.

    Simpler processors like that of the Raspberry Pi are immune to these problems. In effect, the lesson is that more complex systems are more difficult to secure.

Published by

Daniel Lemire

A computer science professor at the Université du Québec (TELUQ).

6 thoughts on “Science and Technology links (January 5th, 2018)”

  1. The solar panel energy balance point appears to have tipped around 2010-4, and the early deficit may be ‘paid off’ by 2020: http://pubs.acs.org/doi/abs/10.1021/es3038824

    OK that’s only one paper, but surely better than a single highly tendentious & unsubstantiated tweet (isn’t uncritical quoting of those how ‘fake news’ spreads?)

    (Admittedly Izabella Kaminska’s wider point is clearly correct)

  2. This quote is grossly inaccurate:
    “You still have to burn more energy creating solar panels than solar panels can give back in a useful lifetime.”

    From this Nature article(1), the energy payback times are ‘under 1 year for poly-Si and just over 1 year for mono-Si PV systems currently’.

    From this NREL study(2), in North America, you should be able to get 25 years of energy out of a solar panel.

    Of course, everybody’s an expert on Twitter, so what do I know?

    1. https://www.nature.com/articles/ncomms13728
    2. https://www.nrel.gov/docs/fy12osti/51664.pdf

  3. Tyler Cowan is well known for publishing outlandish claims from the far right, and should always be fact-checked.

    The solar panel myth is a good example. People were debunking it as early as 2008. https://www.scientificamerican.com/article/solar-cells-prove-cleaner-way-to-produce-power/ The tipping point was widely reported in 2013. https://www.theverge.com/2013/4/2/4174204/solar-panels-finally-generate-more-energy-than-they-consume The scientific literature agrees. https://phys.org/news/2016-12-solar-panels-repay-energy-debt.html

  4. Twitter: *You still have to burn more energy creating solar panels than solar panels can give back in a useful lifetime.*

    Cris: *isn’t uncritical quoting of those how ‘fake news’ spreads?*

    Daniel: *Right, except that I also quote Tyler Cowen who provided a lot more than a tweet.*

    The “fake news” that Cris is objecting to is the potential-fact that the energy required to produce a solar panel is greater than the lifetime energy produced by that panel . Although I haven’t researched it, I’d be quite surprised if the claim is true.

    Your post and response imply that Cowen confirms that energy payback is never reached, but I didn’t see anywhere in the linked page where Cowen even discusses the energy payback period of solar panels. Perhaps it was removed?

    If it’s true that solar panels are a net energy sink, this seems like important knowledge to spread. If false, I’d agree with Cris that this fits into the category of “fake news” even if (as Cowen points out) there are many other problems with full adoption of solar.

Leave a Reply

Your email address will not be published. Required fields are marked *

To create code blocks or other preformatted text, indent by four spaces:

    This will be displayed in a monospaced font. The first four 
    spaces will be stripped off, but all other whitespace
    will be preserved.
    
    Markdown is turned off in code blocks:
     [This is not a link](http://example.com)

To create not a block, but an inline code span, use backticks:

Here is some inline `code`.

For more help see http://daringfireball.net/projects/markdown/syntax