Subscribe to our daily newsletter
Learn the Netflix model of high-performing teams
Erin Meyer explains the keeper test and how it can make or break a team.
27 October, 2020
- There are numerous strategies for building and maintaining a high-performing team, but unfortunately they are not plug-and-play. What works for some companies will not necessarily work for others. Erin Meyer, co-author of No Rules Rules: Netflix and the Culture of Reinvention, shares one alternative employed by one of the largest tech and media services companies in the world.
- Instead of the 'Rank and Yank' method once used by GE, Meyer explains how Netflix managers use the 'keeper test' to determine if employees are crucial pieces of the larger team and are worth fighting to keep.
- "An individual performance problem is a systemic problem that impacts the entire team," she says. This is a valuable lesson that could determine whether the team fails or whether an organization advances to the next level.
<p class="media-headline"><br></p><div class="rm-shortcode amazon-assets-widget" data-rm-shortcode-id="dda52f33ef8e3de23c56f6db42799b3c" contenteditable="false">
<a href="https://www.amazon.com/dp/1984877860?tag=bigthink00-20&linkCode=ogi&th=1&psc=1" target="_blank">
<img type="lazy-image" data-runner-src="https://m.media-amazon.com/images/I/51wfordE5WL.jpg" class="amazon-assets-widget__image widget__image">
<div class="amazon-assets-widget__description">
<div class="amazon-assets-widget__title" style="display: block;">No Rules Rules: Netflix and the Culture of Reinvention</div>
<div class="amazon-assets-widget__by-amazon"><!-- <a href="https://www.amazon.com/dp/1984877860?tag=bigthink00-20&linkCode=ogi&th=1&psc=1" target="_blank">by now at amazone</a> --></div>
<div class="amazon-assets-widget__list-price"><span class="grey">List Price: </span><span class="list-price">$21.06</span></div>
<div class="amazon-assets-widget__new-price"><span class="grey">New From: </span><span class="new-price">$21.06</span> <span class="grey">in Stock</span></div>
<div class="amazon-assets-widget__used-price"><span class="grey">Used From: </span><span class="used-price">$21.08</span> <span class="grey">in Stock</span></div>
</div>
</a>
</div>
<p class="media-headline"><br></p>
<div></div><ul class="ee-ul"></ul>
Keep reading
Show less
Why aligning AI to our values may be harder than we think
Can we stop a rogue AI by teaching it ethics? That might be easier said than done.
19 October, 2020
Eerie looking supercomputer.
Credit: STR/JIJI PRESS/AFP via Getty Images
- One way we might prevent AI from going rogue is by teaching our machines ethics so they don't cause problems.
- The questions of what we should, or even can, teach computers remains unknown.
- How we pick the values artificial intelligence follows might be the most important thing.
<p><br></p>
<p> Plenty of scientists, philosophers, and science fiction writers have wondered how to keep a potential super-human AI from destroying us all. While the obvious answer of "unplug it if it tries to kill you" has many supporters (and it <a href="https://www.youtube.com/watch?v=c8N72t7aScY" rel="noopener noreferrer" target="_blank">worked on </a><a href="https://www.youtube.com/watch?v=c8N72t7aScY" rel="noopener noreferrer" target="_blank">the </a><a href="https://www.youtube.com/watch?v=c8N72t7aScY" rel="noopener noreferrer" target="_blank">HAL </a><a href="https://www.youtube.com/watch?v=c8N72t7aScY" rel="noopener noreferrer" target="_blank">9000</a>), it isn't too difficult to imagine that a sufficiently advanced machine would be able to prevent you from doing that. Alternatively, a very powerful AI might be able to make decisions too rapidly for humans to review for ethical correctness or to correct for the damage they cause. <br> </p><p>The issue of keeping a potentially super-human AI from going rogue and hurting people is called the "control problem," and there are many potential solutions to it. One of the more frequently discussed is "<a href="https://plato.stanford.edu/entries/ethics-ai/#Sing" rel="noopener noreferrer" target="_blank">alignment</a>" and involves syncing AI to human values, goals, and ethical standards. The idea is that an artificial intelligence designed with the proper moral system wouldn't act in a way that is detrimental to human beings in the first place. <br> <br> However, with this solution, the devil is in the details. What kind of ethics should we teach the machine, what kind of ethics <em>can</em> we make a machine follow, and who gets to answer those questions? </p><p><br> Iason Gabriel considers these questions in his new essay, "<a href="https://link.springer.com/article/10.1007/s11023-020-09539-2" rel="noopener noreferrer" target="_blank">Artificial Intelligence, Values, and Alignment.</a>" He addresses those problems while pointing out that answering them definitively is more complicated than it seems. </p>
What effect does how we build the machine have on what ethics the machine can follow?
<iframe width="730" height="430" src="https://www.youtube.com/embed/IHE63fxpHCg" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe><p> <strong><br> </strong>Humans are really good at explaining ethical problems and discussing potential solutions. Some of us are very good at teaching entire systems of ethics to other people. However, we tend to do this using language rather than code. We also teach people with learning capabilities similar to us rather than to a machine with different abilities. Shifting from people to machines may introduce some limitations. <br> <br> Many different methods of machine learning could be applied to ethical theory. The trouble is, they may prove to be very capable of absorbing one moral stance and utterly incapable of handling another. </p><p>Reinforcement learning (RL) is a way to teach a machine to do something by having it maximize a reward signal. Through trial and error, the machine is eventually able to learn how to get as much reward as possible efficiently. With its built-in tendency to maximize what is defined as good, this system clearly lends itself to utilitarianism, with its goal of maximizing the total happiness, and other consequentialist ethical systems. How to use it to effectively teach a different ethical system remains unknown. <br> <br> Alternatively, apprenticeship or imitation learning allows a programmer to give a computer a long list of data or an exemplar to observe and allow the machine to infer values and preferences from it. Thinkers concerned with the alignment problem often argue that this could teach a machine our preferences and values through action rather than idealized language. It would just require us to show the machine a moral exemplar and tell it to copy what they do. The idea has more than a few similarities to <a href="https://bigthink.com/scotty-hendricks/virtue-ethics-the-moral-system-you-have-never-heard-of-but-have-probably-used" target="_self">virtue ethics</a>. </p><p>The problem of who is a moral exemplar for other people remains unsolved, and who, if anybody, we should have computers try to emulate is equally up for debate. <br> <br> At the same time, there are some moral theories that we don't know how to teach to machines. Deontological theories, known for creating universal rules to stick to all the time, typically rely on a moral agent to apply reason to the situation they find themselves in along particular lines. No machine in existence is currently able to do that. Even the more limited idea of rights, and the concept that they should not be violated no matter what any optimization tendency says, might prove challenging to code into a machine, given how specific and clearly defined you'd have to make these rights.</p><p>After discussing these problems, Gabriel notes that:<br> <br> "In the light of these considerations, it seems possible that the methods we use to build artificial agents may influence the kind of values or principles we are able encode."<br> <br> This is a very real problem. After all, if you have a super AI, wouldn't you want to teach it ethics with the learning technique best suited for how you built it? What do you do if that technique can't teach it anything besides utilitarianism very well but you've decided virtue ethics is the right way to go? </p><iframe width="730" height="430" src="https://www.youtube.com/embed/91TRVubKcEM" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe><p>If philosophers can't agree on how people should act, how are we going to figure out how a hyper-intelligent computer should function?</p><p> The important thing might not be to program a machine with the one true ethical theory, but rather to make sure that it is aligned with values and behaviors that everybody can agree to. Gabriel puts forth several ideas on how to decide what values AI should follow. </p><p> A set of values could be found through consensus, he argues. There is a fair amount of overlap in human rights theory among a cross-section of African, Western, Islamic, and Chinese philosophy. A scheme of values, with notions like "all humans have the right to not be harmed, no matter how much economic gain might result from harming them," could be devised and endorsed by large numbers of people from all cultures. </p><p> Alternatively, philosophers might use the "Veil of Ignorance," a thought experiment where people are asked to find principles of justice that they would support if they didn't know what their self-interests and societal status would be in a world that followed those principles, to find values for an AI to follow. The values they select would, presumably, be ones that would protect everyone from any mischief the AI could cause and would assure its benefits would reach everyone. <br> <br> Lastly, we could vote on the values. Instead of figuring out what people would endorse under certain circumstances or based on the philosophies they already subscribe to, people could just vote on a set of values they want any super AI to be bound to. <br> <br> All of these ideas are also burdened by the present lack of a super AI. There isn't a consensus opinion on AI ethics yet, and the current debate hasn't been as cosmopolitan as it would need to be. The thinkers behind the Veil of Ignorance would need to know the features of the AI they are planning for when coming up with a scheme of values, as they would be unlikely to choose a value set that an AI wasn't designed to process effectively. A democratic system faces tremendous difficulties in assuring a just and legitimate "election" for values that everybody can agree on was done correctly. </p><p>Despite these limitations, we will need an answer to this question sooner rather than later; coming up with what values we should tie an AI to is something you want to do <em>before</em> you have a supercomputer that could cause tremendous harm if it doesn't have some variation of a moral compass to guide it. <br> <br> While artificial intelligence powerful enough to operate outside of human control is still a long way off, the problem of how to keep them in line when they do arrive is still an important one. Aligning such machines with human values and interests through ethics is one possible way of doing so, but the problem of what those values should be, how to teach them to a machine, and who gets to decide the answers to those problems remains unsolved. </p>
Keep reading
Show less
Electricity and fear: The trouble with nuclear energy
Although everyone knows that coal-based energy is a thing of the past, declarations about nuclear power plants somehow do not want to enter into force.
17 October, 2020
Photo by Viktor Kiryanov on Unsplash
No other power-generating device raises as much concern as the nuclear reactor. Because of this, until recently the future of the entire energy sector has been determined by its past.
<p>On the eve of the pandemic, the European energy sector found itself at a crossroads, somewhere between Great Britain, Germany and Poland. Five years ago, across the English Channel, the then Prime Minister David Cameron announced an ambitious program to build 12 new nuclear power plants with a total capacity of 16 GW. While developing renewable energy resources, they would allow the United Kingdom to reduce carbon dioxide emissions from the energy sector to almost zero. Soon after, Cameron came up with the idea of a referendum on leaving the EU – and Brexit reset all long-term British plans. However, the British are already producing electricity in a very sustainable way. Almost 38% comes from renewable sources, about 20% from nuclear power plants, while the remainder is provided by gas-powered plants, the only ones that emit CO2.</p><p>Meanwhile in Germany, the aversion towards nuclear energy has been growing for years. Finally, following the Fukushima disaster in March 2011, chancellor Angela Merkel announced that all nuclear power plants would be shut down by 2022. For the first few years, the great <em>Energiewende</em> (energy transformation) plan seemed to be going well. Thanks to subsidies and increased electricity prices for individual customers, the intensive development of wind farms and solar power plants continued. However, no technological solution has been found to overcome the main weakness of renewable energy sources: plants running on renewable sources work on average for 20-30% of the day and remain completely dependent on whether the wind blows or the sun shines. Because of this, they are not able to handle energy peaks. In turn, when a gale comes, suddenly there is a network overload due to excess power. In both these extreme cases, the entire country is at risk of blackout, and the risk of collapsing energy supplies increases significantly when more than 30% is obtained from renewable sources. Safety requires the maintenance of traditional power plants, which due to their flexibility, stabilize the entire system.</p><p>In Germany, as subsequent nuclear reactors began to shut down, lignite-fired power plants started to play a key role. Unlike nuclear plants, they devastate the natural environment not only due to CO2 emissions, but also the need to expand opencast mines. A huge wave of criticism from environmentalists and Berlin's goal to lead by example in the fight against global warming have brought an adjustment in strategy. Today, coal-fired power plants are being replaced by gas-fired ones that emit one-third less carbon dioxide. Russia will provide fuel for them via the Nord Stream and Nord Stream 2 gas pipelines. However, withdrawal from the decommissioning of nuclear power plants is now out of the question.</p>
<p>In turn, the development of renewable energy in Poland is suffering, despite the construction of one or more nuclear power plants having been announced two decades ago. Before the pandemic, the government envoy for strategic energy infrastructure Piotr Naimski claimed that by the end of 2045 as many as six nuclear reactors with a total capacity of 6 GW would be built. Although everyone knows that coal-based energy is a thing of the past, declarations about nuclear power plants somehow do not want to enter into force. And this is a very complicated undertaking, during which any disregard of security standards can awaken demons from the past.</p><h3>A pile of trouble</h3><p>"In fifteen years, nuclear power will provide electricity too cheap to measure its consumption," the head of the American Atomic Energy Commission, Lewis Strauss, prophesied in 1954. By the end of that decade, energy corporations had overcome technological barriers. "Westinghouse has perfected the PWR reactor, the water-pressure reactor, and GE [General Electric] the BWR reactor, boiling water reactor," explains Daniel Yergin in the <em>The Quest: In Search of Energy</em>. These two types of first generation reactors have spread throughout the world. By 1970, 15 nuclear power plants had launched in 62 countries and the construction of a further 89 had begun. Most of them were located in the US, USSR, UK, France, Japan and West Germany. Three years later, the first oil crisis erupted and it seemed certain that highly developed countries would base their future on nuclear power plants. However, the first problems began to emerge.</p><p>The first generation, 1000 MW water-pressure reactor generated as much as 20 tons of radioactive waste annually. Initially, the Americans placed it into metal containers and buried it in the ocean. The Soviets did the same. Protests by environmental organizations led to containers with a guarantee of durability of a thousand years starting to be buried in the Nevada desert – ignoring the fact that the half-life of plutonium-239 is about 24,400 years. In other countries, old mines were used as waste dumps. The French coped with this problem exemplarily by building a plant at La Hague specializing in the recovery of radioactive uranium and plutonium from waste. Later, these elements are enriched and sold to energy companies. During the 1980s, many countries – including Japan, West Germany, Belgium and Switzerland – began to use the services of the French.</p><p>In addition to waste, investment costs have become an equally large problem. "Emerging ecological movements, especially anti-nuclear ones, forced additional reviews and changes. It was necessary to thicken the concrete walls, and remove pipeline installations and rework them. Power plants had to be redesigned, even several times during construction," emphasizes Yergin. He writes: "Power plants also became more expensive because of inflation and later, the high interest rates on loans. Instead of six years, construction took ten; it also cost money. The power plants, which were to cost $200million, ultimately cost $2billion." Later, they produced the cheapest electricity on the market, but gigantic expenses had to be included in its price. While the French model handles waste well, investment costs remain the Achilles' heel of nuclear energy to this day, even if they are less important than the media and public fear.</p>
<h3>Awaiting the apocalypse</h3><p>"There is nothing in the laws of nature that stops us from building better nuclear power plants. We are stopped by a deep justified public distrust. The public distrusts the experts because they claimed to be infallible," writes Freeman Dyson, a physicist who participated in the construction of the first reactors, in the book <em>Imagined Worlds</em>. The distrust of nuclear energy emerged gradually. In the 1960s, everyone remembered the fate of Hiroshima and Nagasaki, but the fear of radioactive radiation had not yet paralysed ordinary people. Experts managed to convince Western societies that the nuclear power plant hardly differed from the coal-fired power plant. All it needs is access to a lot more coolant for the reactor, preferably a huge water tank.</p><p>The sense of security began to fade not because of a failure, but catastrophic scenarios loved by the press, especially in West Germany. In October 1975, <em>Der Spiegel</em> very vividly presented to readers what would happen if the reactor at a power plant built near Ludwigshafen overheated. "The molten reactor core will penetrate the surrounding protective structures. It will sink into the ground at a speed of two to four meters per hour. The amount of radiation emitted would correspond to the radiation of a thousand bombs such as the one dropped on Hiroshima," the newspaper forecasted, estimating the number of victims at 100,000 killed immediately and about 1.6 million "dying slowly" due to radiation sickness. Such apocalyptic visions interested Hollywood, resulting in the neo-thriller entitled <em>The China Syndrome</em>. In specialist jargon, this name means the severe meltdown of the core components of the reactor.</p><p>Lo and behold, two weeks after the film's release, on 28th March 1979, there was a failure at the Three Mile Island nuclear power plant located on an artificial island. Pipes supplying coolant to the reactor burst when the back-up cooling system was disconnected for inspection. The reactor had warmed up, but the safety measures worked. Each reactor is managed using control rods. They are made of alloys that absorb neutrons. Sliding the control rods in between the fuel rods slows down the chain reaction. Pulling them out has the opposite effect. When the reactor overheats, all control rods fall into the core, quenching the reaction.</p><p>This happened at Three Mile Island. However, due to the pipes bursting, water poured out onto the reactor jacket and immediately evaporated, forming a mixture of oxygen and hydrogen under the dome of the power block. One spark could have blown up the power plant. The following day, technicians pumped off hazardous, radioactive gases outside. The residents of nearby Harrisburg panicked. About 80,000 people attempted to escape the city in cars. The US energy minister James Schlesinger's assurances that the radiation only increased by around 0.03 rem and would not hurt anyone fell on deaf ears. Those who have seen <em>The China Syndrome</em> knew better. It wasn't until five days later, when President Jimmy Carter personally visited Three Mile Island and in the presence of TV cameras toured the area, that the panic was subjugated. However, the misfortunes of nuclear power plants were only just beginning.</p>
<h3>The weakest link</h3><p>The owners of the plant, the Westinghouse group, largely caused the Three Mile Island disaster. The power plant was built in a rush to make it operational before 30th December 1978, in order for the company to gain a $40 million tax break. After launching the reactor, it turned out that the coolant supply pipes were leaking. At that point, the management ordered temporary sealing of leaks, after which the test of the emergency cooling system was performed, starting with its shutdown. This was done on the assumption that the main pipes would still last a little longer. "The accident was caused by a series of relatively small equipment failures followed by operator error," the head of the commission investigating the causes of the disaster, Admiral Hyman G. Rickover, wrote in his report. Fortunately, none of the Westinghouse executives were so thoughtless as to deactivate the other safeguards. Seven years later, it turned out that even such recklessness is possible.</p><p>On the night of 26th April 1986, the management of the <a href="https://przekroj.pl/en/culture/comrade-batman-victor-martinovich" target="_blank" rel="noopener noreferrer">Chernobyl power plant </a>began to experiment with manual control of the reactor in block 4. For complete freedom, all automatic security systems were turned off. During the experiments, the stack heated up rapidly, and the control rods blocked by the staff did not automatically quench the chain reaction. Then the pipes supplying water to the cooling system burst. As in Three Mile Island, the water evaporated by the hot reactor turned into hydrogen and oxygen. The explosion of this mixture tore the dome and threw a 500-ton piece of concrete into the air, which a moment later fell into the reactor, breaking it completely. 50 tons of fuel escaped outside and the core melted. Vast areas of northern Ukraine and Belarus became contaminated due to the radioactive cloud. 50,000 residents of the nearby town of Pripyat and surrounding villages were evacuated.</p><p>As a result of the disaster, 31 people lost their lives (mainly irradiated firefighters). UNSCEAR (UN Scientific Committee on Effects of Atomic Radiation) found that there were many more casualties: a 2000 report found that of about 600 employees of the power plant and firefighters, 237 were diagnosed with radiation sickness symptoms. Of these, 28 people died. According to the report, epidemiologists have not observed an increase in the incidence of cancer in the most contaminated areas, except for higher than average rates of thyroid cancer. No genetic defects were found in the offspring of irradiated persons.</p>
<h3>The wave</h3><p>A quarter of a century later, the 'Chinese syndrome' became Japanese. Two oil crises in the 1970s encouraged the government of Japan to finance the construction of 50 nuclear reactors. They guaranteed energy security for the state. However, haste made them forget about their side effects in a country where earthquakes happen regularly. The Fukushima reactor was built right on the seafront. When massive shocks (9 on the Richter scale) came on <a href="https://przekroj.pl/en/literature/what-lies-under-the-water-katarzyna-boni" target="_blank" rel="noopener noreferrer">11th March 2011</a>, the security systems functioned properly. The reactors were automatically quenched and the cooling system switched to the emergency power supply. Nothing bad would have happened if it weren't for the sea. Tectonic shocks caused a tsunami wave of 15-metre heights, and the breakwater was only six-metres high. Huge amounts of water flooded the power plant. The power generators went down and the reactor core suddenly stopped being cooled. Then the water evaporated and the hydroxide mixture exploded.</p><p>About 10 times less radioactive substance escaped outside than in Chernobyl, and no-one was killed during the event. The first person irradiated as a result of the disaster's aftermath did not die until September 2018. Yet again, however, a wave of fear swept through the entire world.</p>
<h3>The sum of fears</h3><p>The disaster in Fukushima was a strong blow to the nuclear energy sector – which even without it, suffered bad press – and led to public trepidation, even though by the mid-1980s the number of reactors operating worldwide had reached 430 and stopped growing. New ones were still being built in France, Japan, the USSR (later, Russia), South Korea and China, but elsewhere they were gradually dismantled. The only country that had based their entire energy system on nuclear power plants was France, where they produce over 80% of electricity. Finland is also focusing on the development of nuclear energy. Two nuclear power plants currently generate around 30% of the country's energy, and once the third one is built, this will reach 60% (the rest is to come from renewable sources).</p><p>Most countries, however, still recognize the nuclear industry as a dead end. The emergence of much better third generation reactors that use less uranium, while reducing the amount of waste, did not change that. Developed by two companies – the French Framatome and the German Siemens – the EPR (European Pressurized Reactor) has a quadruple safety system and reinforcement that can withstand even the impact of an aircraft crash. In turn, the ESBWR (Economic Simplified Boiling Water Reactor) by GE Hitachi, apart from showing similar resistance, requires minimal amounts of coolant and discharges excess heat directly into the atmosphere.</p><p>There are more innovative constructions, but they have started to generate interest only recently, thanks to the rapid development of Asian countries, and thus an increase in demand for cheap electricity. A nuclear power plant uses roughly 30-50 tons of uranium per year. At a market price of around $55 per kilogram, a cost of fuel of around $2.5 million a year is very cheap – 100 times cheaper than the cost of fuel for a coal-fired power plant fuel. It is estimated that known uranium deposits will last for about 300 years. At the same time, as with crude oil, this deadline may prove to be much more distant, since no new ones have been sought for years. Therefore, it should not come as a surprise that in April 2019 China presented a plan for the vast expansion of its nuclear energy sector. While today the total capacity of Chinese nuclear power plants is about 42 GW, it will exceed 100 GW in 100 years. Then, the People's Republic of China will overtake the US in this field. South Korea has presented slightly less ambitious goals, announcing an increase in nuclear power by one-third.</p><p>And what path will the European Union take? The fight against CO2 emissions determines the direction of its energy policy, and renewable energy sources are a priority. However, to fully base their economy on them, efficient energy storage is necessary – methods capable of accumulating electricity at times of overproduction and releasing it in the absence of sun and wind. Even lithium-ion cells cannot fully cope with this task. Attempts are being made at avoiding the lack of this element by designing self-sufficient buildings that draw energy from solar batteries and heat pumps. However, in the scale of cities and entire countries, large power plants cannot be replaced, and the only ones that do not emit carbon dioxide are nuclear power plants. This fact means that even in Europe, their slow renaissance continues. For now, countries on the outskirts of the EU (Finland, Hungary, Lithuania, the Czech Republic and Slovakia) are modernizing old plants or building new ones. In just one year, the construction of over 60 new reactors began.</p><p>Despite public resentment, more investments will begin soon. Right now, fear of the 'China syndrome' is weaker than fear of the effects of global warming and sudden energy shortages and blackouts.</p><p><em>Translated from </em><a href="https://przekroj.pl/nauka/prad-i-lek-andrzej-krajewski" target="_blank" rel="noopener noreferrer"><em>the Polish</em></a><em> by Joanna Figiel</em></p>Reprinted with permission of <a href="https://przekroj.pl/en/" target="_blank" rel="noopener noreferrer" style="">Przekrój</a>. Read the <a href="https://przekroj.pl/en/science/electricity-and-fear-andrzej-krajewski" target="_blank">original article</a>.
<p><br/></p>
Keep reading
Show less
Solar PV
Credit: beysim via AdobeStock
Technology & Innovation
- The International Energy Agency is an intergovernmental organization that advises member nations on issues related to energy and the environment.
- In its annual report, the IEA reported that the cost of solar is dropping more rapidly than previously thought, providing some parts of the world with historically cheap electricity.
- The IEA predicted that, over the next decade, renewables will meet 80 percent of global electricity demand growth, while the demand for oil will peak.
<p>Electricity from solar energy is the cheapest it's ever been, thanks largely to technological improvements and policies that reduce the risk of investing in renewable energy.</p><p>That's one of the key takeaways from the <a href="https://webstore.iea.org/world-energy-outlook-2020" target="_blank">World Energy Outlook 2020</a> report, which is published annually by the International Energy Agency (IEA), an intergovernmental organization that advises member nations on energy security, economic development and environmental protection.</p><p style="margin-left: 20px;">"I see solar becoming the new king of the world's electricity markets," Dr. Fatih Birol, the IEA Executive Director, said in a <a href="https://www.iea.org/news/world-energy-outlook-2020-shows-how-the-response-to-the-covid-crisis-can-reshape-the-future-of-energy?utm_content=buffer89991&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer" target="_blank">statement</a>. "Based on today's policy settings, it is on track to set new records for deployment every year after 2022."</p>
<div id="9f47d" class="rm-shortcode" data-rm-shortcode-id="71012b8521d232a09588028335fe0de2"><blockquote class="twitter-tweet twitter-custom-tweet" data-twitter-tweet-id="1315865211107237889" data-partner="rebelmouse"><div style="margin:1em 0">The World Energy Outlook 2020 is out!
It shows how the Covid crisis has brought deep disruption & uncertainty to t… https://t.co/if9yt1mHZc</div> — Fatih Birol (@Fatih Birol)<a href="https://twitter.com/IEABirol/statuses/1315865211107237889">1602561688.0</a></blockquote></div><p>The IEA's 2020 report notes that the COVID-19 pandemic has made it hard to predict the future of global energy demand, but it laid out several scenarios of what may happen over the next few years. In the main scenario, the international community is projected to generate 43 percent more solar power than the IEA had predicted in 2018.</p><p>That increase, as Carbon Brief first <a href="https://www.carbonbrief.org/solar-is-now-cheapest-electricity-in-history-confirms-iea" target="_blank">reported</a>, is partly due to new analyses showing the cost of solar power to be 20 to 50 percent cheaper than the IEA thought in 2018. </p><p>Globally, the average cost of electricity from large-scale solar photovoltaic projects has dropped by magnitudes over the past decade, from 38 cents per kilowatt-hour in 2010 to 6.8 cents per kilowatt-hour in 2019, according to the <a href="https://www.irena.org/newsroom/pressreleases/2020/Jun/Renewables-Increasingly-Beat-Even-Cheapest-Coal-Competitors-on-Cost" target="_blank">International Renewable Energy Agency</a>.</p><p style="margin-left: 20px;"><span></span>"Solar PV is now consistently cheaper than new coal- or gas-fired power plants in most countries, and solar projects now offer some of the lowest cost electricity ever seen," the IEA wrote in a <a href="https://www.iea.org/news/world-energy-outlook-2020-shows-how-the-response-to-the-covid-crisis-can-reshape-the-future-of-energy?utm_content=buffer89991&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer" target="_blank">press release</a>. "In the [main scenario], renewables meet 80% of global electricity demand growth over the next decade. Hydropower remains the largest renewable source, but solar is the main source of growth, followed by onshore and offshore wind."</p><p>Solar energy and other renewables also pose a major threat to coal. The IEA report notes that the COVID-19 pandemic has catalyzed a "structural fall in global coal demand," and that renewables and "cheap natural gas and coal phase-out policies, means that coal demand in advanced economies drops by almost half to 2030." </p><p>What's more, India, the world's second-largest coal producer, is projected to build 86 percent less new coal power capacity than the IEA predicted in 2019.</p>
<div id="8618b" class="rm-shortcode" data-rm-shortcode-id="32c4298e9008bdf071c8d149dd0a05f5"><blockquote class="twitter-tweet twitter-custom-tweet" data-twitter-tweet-id="1316682422268891137" data-partner="rebelmouse"><div style="margin:1em 0">Here's a remarkable detail buried in @IEA #WEO20
India will build 86% less new coal power capacity than expected l… https://t.co/oBKGtDRzpT</div> — Simon Evans (@Simon Evans)<a href="https://twitter.com/DrSimEvans/statuses/1316682422268891137">1602756526.0</a></blockquote></div><p>While it's relatively easy to predict the near-term rise of solar and fall of coal, the future of oil remains more uncertain. The report notes that, without additional policy pushes, it's too soon to predict a rapid decline in oil demand and production. Yet, some industry experts do expect to see oil demand level out, if not rapidly decline, over the next decade.</p><p style="margin-left: 20px;">"The era of global oil demand growth will come to an end in the next decade," Dr Birol said in a <a href="https://www.iea.org/news/world-energy-outlook-2020-shows-how-the-response-to-the-covid-crisis-can-reshape-the-future-of-energy?utm_content=buffer89991&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer" target="_blank">press release</a>. "Based on today's policy settings, a global economic rebound would soon push oil demand back to pre-crisis levels."</p><p>One scenario in the report describes a pathway that would see the global community achieve net-zero CO2 emissions by 2050 and limit global warming to 1.5 degrees Celsius, which is the target of the United Nations Framework Convention on Climate Change.</p>
<div id="0c02d" class="rm-shortcode" data-rm-shortcode-id="049f0c388111e6bbd2a9f21a447e305e"><blockquote class="twitter-tweet twitter-custom-tweet" data-twitter-tweet-id="1316650942352195586" data-partner="rebelmouse"><div style="margin:1em 0">THREAD: The @IEA now has an aggressive 1.5°C scenario, reaching net-zero by 2050.
It builds on the Sustainable Dev… https://t.co/Fjc86yHX3Q</div> — Glen Peters (@Glen Peters)<a href="https://twitter.com/Peters_Glen/statuses/1316650942352195586">1602749021.0</a></blockquote></div><p>To hit that ambitious goal, renewables would have to meet roughly 75 percent of global electricity demand in 2030, compared to 40 percent today, and electric vehicles would have to account for 50 percent of all passenger cars sold worldwide by 2030.</p><p>On the individual level, behavioral changes would also play a key role in this scenario. The report notes that up one-third of cuts to CO2 emissions would come from individuals doing things like:</p><ul><li>Working from home more often</li><li>Line-drying laundry</li><li>Driving more slowly</li><li>Reducing use of air conditioning</li><li>Flying less</li></ul>
Keep reading
Show less
Viewing nature in VR or on television boosts wellbeing, study finds
"Interacting" with nature through virtual reality applications had especially strong benefits, according to the study.
15 October, 2020
Hawksbill Sea Turtle in Indian ocean
Credit: Jag_cz via Adobe Stock
Mind & Brain
- Previous studies have shown that spending time in nature can lead to a variety of mental and physical health benefits.
- The new study involved exposing people to a high-definition nature program through one of three mediums: TV, VR and interactive VR.
- The results suggest that nature programs may be an easy and effective way to give people a "dose" of nature, which may be especially helpful during pandemic lockdowns.
<p>Spending time in nature can bring you <a href="https://bigthink.com/surprising-science/ecotherapy-benefits?rebelltitem=1#rebelltitem1" target="_self">well-established health benefits</a>, from lowered anxiety and depression, to reduced blood pressure and a stronger immune system. But how nature produces these effects remains unclear. Is it the awe you feel hiking through a centuries-old forest? Time spent away from screens? The physical exercise?</p><p>Recent research adds a new dimension to scientists' understanding of how nature impacts wellbeing. <a href="https://www.exeter.ac.uk/news/research/title_821333_en.html" target="_blank">The study</a>, published in the Journal of Environmental Psychology, found that people who watched high-definition nature programs on TV or in virtual reality reported lower levels of boredom and other negative emotions. </p><p>For the study, researchers at the University of Exeter asked 96 participants to watch a video of a person describing their job at an office supply company. This was done to induce boredom. </p><p>The participants then watched or interacted with a nature program about a coral reef (which includes several scenes from the BBC's "Blue Planet II" series) under one of three randomly assigned conditions: </p><ul><li>2D video viewed on a high-definition TV screen</li><li>360-degree VR, viewed via a head mounted display (HMD)</li><li>Interactive computer-generated VR (CG-VR), also viewed via a HMD and interacted with using a hand-held controller</li></ul>
<img type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8yNDUyMTIyMC9vcmlnaW4ucG5nIiwiZXhwaXJlc19hdCI6MTYyMTk0MTA0OX0.WD4pU0_oQe_93_n0nDzT0GwawBQuhcpdNgpuwrupKJE/img.png?width=980" id="9722b" class="rm-shortcode" data-rm-shortcode-id="e0003021b52fc8d173869040d1b2eba5" data-rm-shortcode-name="rebelmouse-image" />
Credit: Yeo et al.
<p>The results showed that watching the nature program under all three conditions lowered negative affect, including emotions like boredom and sadness. But only the group who experienced the program in interactive VR reported a boost in mood, and feelings of being more connected to nature.</p><p style="margin-left: 20px;">"Our results show that simply watching nature on TV can help to lift people's mood and combat boredom," lead researcher Nicky Yeo <a href="https://www.exeter.ac.uk/news/research/title_821333_en.html" target="_blank">told</a> University of Exeter News. "With people around the world facing limited access to outdoor environments because of COVID-19 quarantines, this study suggests that nature programmes might offer an accessible way for populations to benefit from a 'dose' of digital nature."</p>Helping those without access to nature
<p>"Dose" is probably a keyword: The researchers didn't compare the benefits of experiencing nature via TV or VR to experiencing it in person. But even beyond the pandemic, the findings suggest that experiencing nature via virtual reality could help people improve their mental wellbeing — a tool that could prove especially useful for people who don't live near natural environments.</p><p style="margin-left: 20px;">"Virtual reality could help us to boost the wellbeing of people who can't readily access the natural world, such as those in hospital or in long term care," co-author Mathew White told University of Exeter News. "But it might also help to encourage a deeper connection to nature in healthy populations, a mechanism which can foster more pro-environmental behaviours and prompt people to protect and preserve nature in the real world."</p><span style="display:block;position:relative;padding-top:56.25%;" class="rm-shortcode" data-rm-shortcode-id="7dbb34dc2cdb359617d23eeb4f554801"><iframe type="lazy-iframe" data-runner-src="https://www.youtube.com/embed/7AkbUfZjS5k?rel=0" width="100%" height="auto" frameborder="0" scrolling="no" style="position:absolute;top:0;left:0;width:100%;height:100%;"></iframe></span><p>High-end VR headsets remain prohibitively costly for many consumers. One of the cheapest models, the Oculus Quest 2, costs $300, while more advanced headsets can run upward of $1,000. Still, you can buy barebones devices, like Google Cardboard, for about $10. These don't enable you to engage in fully interactive VR applications, but you could use them to view 360-degree <a href="https://youtu.be/mBWj443FNGs" target="_blank">virtual reality nature videos on YouTube</a>.<br></p>
<p class=""><br></p>
Keep reading
Show less
Six ways machine learning threatens social justice
Machine learning is a powerful and imperfect tool that should not go unmonitored.
15 October, 2020
Credit: Monopoly919 on Adobe Stock
Technology & Innovation
- When you harness the power and potential of machine learning, there are also some drastic downsides that you've got to manage.
- Deploying machine learning, you face the risk that it be discriminatory, biased, inequitable, exploitative, or opaque.
- In this article, I cover six ways that machine learning threatens social justice and reach an incisive conclusion: The remedy is to take on machine learning standardization as a form of social activism.
<p>When you harness the power and potential of machine learning, there are also some drastic downsides that you've got to manage. Deploying machine learning, you face the risk that it be discriminatory, biased, inequitable, exploitative, or opaque. In this article, I cover six ways that machine learning threatens social justice and reach an incisive conclusion: The remedy is to take on machine learning standardization as a form of social activism.</p><p>When you use machine learning, you aren't just optimizing models and streamlining business.<strong> You're governing. </strong>In essence, the models embody policies that control access to opportunities and resources for many people. They drive consequential decisions as to whom to investigate, incarcerate, set up on a date, or medicate – or to whom to grant a loan, insurance coverage, housing, or a job.<br></p><p>For the same reason that machine learning is valuable—that it drives operational decisions more effectively—it also wields power in the impact it has on millions of individuals' lives. Threats to social justice arise when that impact is detrimental, when models systematically limit the opportunities of underprivileged or protected groups.</p>
Here are six ways machine learning threatens social justice
<img type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8yNDUyMDgxNC9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTY0MzM0NjgxOH0.zHvEEsYGbNA-lnkq4nss7vwVkZlrKkuKf0XASf7A7Jg/img.jpg?width=980" id="05f07" class="rm-shortcode" data-rm-shortcode-id="a7089b6621166f5a2df77d975f8b9f74" data-rm-shortcode-name="rebelmouse-image" />Credit: metamorworks via Shutterstock
<p><strong></strong><strong>1) </strong><strong>Blatantly discriminatory models</strong> are predictive models that base decisions partly or entirely on a protected class. Protected classes include race, religion, national origin, gender, gender identity, sexual orientation, pregnancy, and disability status. By taking one of these characteristics as an input, the model's outputs – and the decisions driven by the model – are based at least in part on membership in a protected class. Although models rarely do so directly, there is <a href="https://www.youtube.com/watch?v=eSlzy1x6Fy0" target="_blank">precedent</a> and <a href="https://www.youtube.com/watch?v=wfpNN8ASIq4" target="_blank">support</a> for doing so.</p><p>This would mean that a model could explicitly hinder, for example, black defendants for being black. So, imagine sitting across from a person being evaluated for a job, a loan, or even parole. When they ask you how the decision process works, you inform them, "For one thing, our algorithm penalized your score by seven points because you're black." This may sound shocking and sensationalistic, but I'm only literally describing what the model would do, mechanically, if race were permitted as a model input. </p><p><strong>2) Machine bias</strong>. Even when protected classes are not provided as a direct model input, we find, in some cases, that model predictions are still inequitable. This is because other variables end up serving as proxies to protected classes. This is <a href="https://coursera.org/share/51350b8fb12a5937bbddc0e53a4f207d" target="_blank" rel="noopener noreferrer">a bit complicated</a>, since it turns out that models that are fair in one sense are unfair in another. </p><p>For example, some crime risk models succeed in flagging both black and white defendants with equal precision – each flag tells the same probabilistic story, regardless of race – and yet the models falsely flag black defendants more often than white ones. A crime-risk model called COMPAS, which is sold to law enforcement across the US, falsely flags white defendants at a rate of 23.5%, and Black defendants at 44.9%. In other words, black defendants who don't deserve it are <a href="https://coursera.org/share/df6e6ba7108980bb7eeae0ba22123ac1" target="_blank" rel="noopener noreferrer">erroneously flagged almost twice as much</a> as white defendants who don't deserve it.</p><p><strong>3) Inferring sensitive attributes</strong>—predicting pregnancy and beyond. Machine learning predicts sensitive information about individuals, such as sexual orientation, whether they're pregnant, whether they'll quit their job, and whether they're going to die. Researchers have shown that it is possible to <a href="https://youtu.be/aNwvXhcq9hk" target="_blank" rel="noopener noreferrer">predict race based on Facebook likes</a>. These predictive models deliver dynamite.</p><p>In a particularly extraordinary case, officials in China use facial recognition to <a href="https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html" target="_blank" rel="noopener noreferrer">identify and track the Uighurs, a minority ethnic group</a> systematically oppressed by the government. This is the first known case of a government using machine learning to profile by ethnicity. One Chinese start-up valued at more than $1 billion said its software could recognize "sensitive groups of people." It's website said, "If originally one Uighur lives in a neighborhood, and within 20 days six Uighurs appear, it immediately sends alarms" to law enforcement.</p><span style="display:block;position:relative;padding-top:56.25%;" class="rm-shortcode" data-rm-shortcode-id="10134f9910c910d2002e2c1984069791"><iframe type="lazy-iframe" data-runner-src="https://www.youtube.com/embed/IHE63fxpHCg?rel=0" width="100%" height="auto" frameborder="0" scrolling="no" style="position:absolute;top:0;left:0;width:100%;height:100%;"></iframe></span><p><strong></strong><strong>4) A lack of transparency.</strong> A computer can keep you in jail, or deny you a job, a loan, insurance coverage, or housing – and yet <a href="https://coursera.org/share/76491423c1a084324962c9ee8abd6b9e" target="_blank">you cannot face your accuser</a>. The predictive models generated by machine learning to drive these weighty decisions are generally kept locked up as a secret, unavailable for audit, inspection, or interrogation. Such models, inaccessible to the public, perpetrate a lack of due process and a lack of accountability.</p><p>Two ethical standards oppose this shrouding of electronically-assisted decisions: 1) <em>model transparency</em>, the standard that predictive models be accessible, inspectable, and understandable. And 2) <em>the right to explanation</em>, the standard that consequential decisions that are driven or informed by a predictive model are always held up to that standard of transparency. Meeting those standards would mean, for example, that a defendant be told which factors contributed to their crime risk score -- which aspects of their background, circumstances, or past behavior caused the defendant to be penalized. This would provide the defendant the opportunity to respond accordingly, establishing context, explanations, or perspective on these factors.</p><p><strong>5) Predatory micro-targeting.</strong> Powerlessness begets powerlessness – and that cycle can magnify for consumers when machine learning increases the efficiency of activities designed to maximize profit for companies. Improving the micro-targeting of marketing and the predictive pricing of insurance and credit can magnify the cycle of poverty. For example, highly-targeted ads are more adept than ever at <a href="https://youtu.be/LeZv2RanMRQ" target="_blank">exploiting vulnerable consumers</a> and separating them from their money.</p><p>And insurance pricing can lead to the same result. With insurance, the name of the game is to charge more for those at higher risk. Left unchecked, this process can quickly slip into predatory pricing. For example, a churn model may find that elderly policyholders don't tend to shop around and defect to better offers, so there's less of an incentive to keep their policy premiums in check. And pricing premiums based on other life factors also contributes to a cycle of poverty. For example, individuals with poor credit ratings are charged more for car insurance. In fact, a low credit score can increase your premium more than an at-fault car accident.</p><p><strong>6) The coded gaze.</strong> If a group of people is underrepresented in the data from which the machine learns, the resulting model won't work as well for members of that group. This results in <a href="https://www.youtube.com/watch?v=UG_X_7g63rY" target="_blank" rel="noopener noreferrer">exclusionary experiences and discriminatory practices</a>. This phenomenon can occur for both facial image processing and <a href="https://www.pnas.org/content/117/14/7684" target="_blank" rel="noopener noreferrer">speech recognition</a>.</p>
Recourse: Establish machine learning standards as a form of social activism
<p>To address these problems, take on machine learning standardization as a form of social activism. We must establish standards that go beyond nice-sounding yet vague platitudes such as "be fair", "avoid bias", and "ensure accountability". Without being precisely defined, these catch phrases are subjective and do little to guide concrete action. Unfortunately, such broad language is fairly common among the principles released by many companies. In so doing, companies protect their public image more than they protect the public.<br></p><p>People involved in initiatives to deploy machine learning have a powerful, influential voice. These relatively small numbers of people mold and set the trajectory for systems that automatically dictate the rights and resources that great numbers of consumers and citizens gain access to.</p><p>Famed machine learning leader and educator Andrew Ng drove it home: "AI is a superpower that enables a small team to affect a huge number of people's lives... Make sure the work you do leaves society better off."</p><p>And Allan Sammy, Director, Data Science and Audit Analytics at Canada Post, clarified the level of responsibility: "A decision made by an organization's analytic model is a decision made by that entity's senior management team."</p><p>Implementing ethical data science is as important as ensuring a self-driving car knows when to put on the breaks.</p><p>Establishing well-formed ethical standards for machine learning will be an intensive, ongoing process. For more, <a href="https://youtu.be/ToSj0ZkJHBQ" target="_blank">watch this short video</a>, in which I provide some specifics meant to kick-start the process.</p><p><em>Eric Siegel, Ph.D., is a leading consultant and former Columbia University professor who makes machine learning understandable and captivating. He is the founder of the long-running <a href="https://www.predictiveanalyticsworld.com/" target="_blank" rel="noopener noreferrer">Predictive Analytics World</a> and the <a href="https://www.deeplearningworld.com/" target="_blank" rel="noopener noreferrer">Deep Learning World</a> conference series and the instructor of the end-to-end, business-oriented Coursera specialization </em><em><a href="http://www.machinelearning.courses/" target="_blank" rel="noopener noreferrer">Machine learning for Everyone</a>. Stay in touch with Eric on Twitter <a href="https://twitter.com/predictanalytic" target="_blank">@predictanalytic.</a></em><a href="http://www.machinelearning.courses/" target="_blank" rel="noopener noreferrer"></a></p>
Keep reading
Show less
Popular
