Sunday, June 16, 2019
Home Blog

When will voice assistants understand what we are saying? Probably quite long.

Humans are amazingly good at dealing with variations in language. Technology cannot quite cope with it. Technology cannot quite cope with it. One could, laboriously, teach different accents to a speech recognition system, but accent variation is just the tip of the iceberg | Aspioneer
orange blue faces of people

Imagine a world in which Siri always understands you, Google Translate works perfectly, and the two of them create something akin to a Doctor Who style translation circuit. Imagine being able to communicate freely wherever you go (not having to mutter in school French to your Parisian waiter). It’s an attractive, but still distant prospect. One of the bottlenecks in moving this reality forward is variation in language, especially spoken language. Technology cannot quite cope with it.

Humans, on the other hand, are amazingly good at dealing with variations in language. We are so good, in fact, that we really take note when things occasionally break down. When I visited New Zealand, I thought for a while that people were calling me “pet”, a Newcastle-like term of endearment. They were, in fact, just saying my name, Pat. My aha moment happened in a coffee shop (“Flat white for pet!” gave me a pause).

This story illustrates how different accents of English have slightly different vowels – a well-known fact. But let’s try to understand what happened when I misheard the Kiwi pronunciation of Pat as pet. There is a certain range of sounds that we associate with vowels, like a or e. These ranges are not absolute. Rather, their boundaries vary, for instance between different accents. When listeners fail to adjust for this, as I did in this case, the mapping of sound to meaning can be distorted.

One could, laboriously, teach different accents to a speech recognition system, but accent variation is just the tip of the iceberg. Vowel sounds can also vary depending on our age, gender, social class, ethnicity, sexual orientation, level of intoxication, how fast we are talking, whom we are talking to, whether or not we are in a noisy environment … the list just goes on, and on.

The crux/crooks of the matter

Consider that a recent study I was involved in showed that even moving house (or not) can affect one’s vowels. Specifically, there is a correlation between how speakers of Northern English pronounce the vowel in words like crux, and how many times they have moved in the last decade. People who have not moved at all are more likely to pronounce crux the same as crooks, which is the traditional Northern English pronunciation. But those who have moved four times or more are more likely to have different vowels in the two words, similarly in the south of England.

There is, of course, nothing about the act of moving that causes this. But moving house multiple times is correlated with other lifestyle factors, for instance interacting with more people, including people with different accents, which might influence the way we speak.

Other sources of variation may have to do with linguistic factors, such as word structure. A striking example comes from pairs of words such as ruler, meaning “measuring device” and ruler, meaning “leader”.

These two words are superficially identical, but they differ at a deeper structural level. A rul-er is someone who rules, just like a sing-er is someone who sings, so we can analyse these words as consisting of two meaningful units. In contrast, ruler meaning “measuring device” cannot be decomposed further.

It turns out that the two meanings of ruler are associated with a different vowel for many speakers of Southern British English, and the difference between the two words has increased in recent years: it is larger for younger speakers than it is for older speakers. So both hidden linguistic structure and speaker age can affect the way we pronounce certain vowels.

End never in sight

This illustrates another important property of language variation: it keeps changing. Language researchers therefore constantly have to review their understanding of variation, which in turn requires continuing to acquire new data, and updating the analysis. The way we do this in linguistics is being revolutionised by new technologies, advances in instrumental data analysis, and the ubiquity of recording equipment (in 2018, 82% of the UK adult population owned a recording device, otherwise known as a smartphone).

Modern day linguistic projects can profit from the technological advancement in various ways. For instance, the English Dialects App collects recordings remotely via smartphones, to build a large and constantly updating corpus of modern day English accents. That corpus is the source of the finding concerning the vowel in crux in Northern English, for example. Accumulating information from this and many other projects allows us to track variation with increased coverage, and to build ever more accurate models predicting the realisation of individual sounds.

Can this newly refined linguistic understanding also improve speech recognition technology? Perhaps, but in order to improve, the technology needs to know a lot more about you.

This article was originally published on The Conversation. Read the original article.

Why the explosive growth of e-commerce might be coming to an end

There is an inflection point approaching, when consumers will either have to pay more for online purchases or end up with fewer products and services to choose from | Aspioneer
Laptop showing sale and consumer ready with card to make purchase

Many people probably assume that online stores are making a fortune, without all the costly bricks and mortar. But the reality is rather different. Many ecommerce activities are, in fact, unprofitable; if people had to pay the true cost of what they bought online, they would probably buy less. In fact, we think there is an inflection point approaching, when consumers will either have to pay more for online purchases or end up with fewer products and services to choose from.

Let’s start with the online retail leviathan Amazon, which chalked up record profits and revenues in 2018. This is great news for Amazon shareholders, but deeper scrutiny reveals a different picture. To begin with, most of the profit was not from retail activities. Amazon Web Services, a cloud-hosting business unrelated to ecommerce, generated more operating income than the company’s entire North American retail operation – and with margins over five times higher.

Even then, this was a much better performance from the retail division than in 2017, when the North American operating income was completely offset by international retail losses. In that year, Amazon’s positive operating income was entirely thanks to the cloud-hosting business.

Profit push

Amazon’s retail improvement in 2018 came on the back of a profitability drive, much of which involved raising the consumer cost of ecommerce. For example, Amazon increased the annual membership cost of priority customer service Prime by 20% to US$119 (£94) in the US, along with comparable rises in other countries.

According to one estimate, this US hike accounted for nearly a third of Amazon North America’s operating margin in 2018. Yet not all of this extra profitability looks sustainable: Amazon is now seeing shrinking growth in Prime membership in North America and declines in some countries as customers at the margin decide to walk away.

Amazon has also been targeting its CRaP products, which stands for “cannot return a profit”. Product lines end up in this category because of small margins or logistical challenges such as their weight or size. Bottled water, fizzy drinks and snack foods are all examples.

Amazon has been pressuring the manufactures of these products to lower sales costs. It’s unlikely that this will succeed on the whole, since in many cases there’s little room for improvement. This will force Amazon to choose between charging more for these products or delisting them, which will translate into higher prices for consumers or a narrower selection on the site.

Not all of Amazon’s initiatives are at the expense of the consumer, it should be said. The company recently reported a 4% drop in the cost of fulfilling orders, mainly because it has been building fewer new warehouses and ramping up throughput at existing sites instead. This is a welcome development for the company, since the costs of both fulfilling orders and shipping increased as a percentage of sales each year between 2010 and 2017.

Within its warehouse network, Amazon handles own-brand goods and those of many of the other vendors who sell via the platform. These vendors have the choice between paying Amazon a premium to completely handle their distribution and pricing, giving them full access to the Prime customer base; or having a looser relationship that can involve paying Amazon or an independent logistics company to use the warehouse network instead.

Amazon has succeeded in growing these different types of looser relationships – they now make up over half of total retail sales. Developing the third-party logistics strand is creating a new revenue stream and lowering working capital, since it means that Amazon covers less of the cost of overall sales fulfilment. This resembles the business model of the Chinese ecommerce giant Alibaba. Yet saving on working capital doesn’t represent an inherent efficiency, since offloading some distribution expenses is likely to eventually be passed on to consumers as higher prices from costs incurred elsewhere.


Major rival Walmart has its own techniques for trying to make online sales more profitable. Its new approach to CRaP products is to hide them from view in Walmart consumer search results, showing as out of stock alongside alternatives that are more profitable to the company.

Interestingly, Walmart is also piloting free next-day deliveries from its stores in the US without customers having to be members of any Prime-equivalent service. The wrinkle is that the offering is limited to only high-volume, higher margin products. In both examples, Walmart is therefore pruning consumer choice in its search for more profitability online.

Walmart is also one of numerous big retailers that offer same-day grocery delivery, but this too is not all it seems.

An experienced grocery retail manager has told us that online grocery is necessary as a marketing loss leader but “impossible” to make money from. Such delivery offers are only possible, he said, because online grocery is just 2% of the overall market, since most consumers don’t buy these products online. A recent study agreed with this thinking, finding that online grocery orders have a negative margin of about 15%. It is reminiscent of that old business joke about losing money on every sale but making it up in volume.

To understand the mindset of retailers like Walmart and other smaller rivals who are not purely online, a supply chain consultant told us last year that they are placing a priority on speed of change before profitability, amid pressure to stay competitive with the likes of Amazon. “It’s a logic of desperation as much as it is of strategy,” he said.

We can see the consequences in an interesting survey which found that in 2017, 61% of supply chain executives reported increasing product lines due to ecommerce, up from 55% in 2013. When asked about the impacts on distribution, 26% said they were implementing smaller, more localised warehouses, up from 20% in 2013. These changes inevitably lead to higher costs, which will again be passed on, at least in part, to the consumer.

Viewed as a whole, the inflection point in online shopping that we mentioned earlier could be getting close. We may have reached peak convenience and cheap prices, and might now be entering a world of more targeted offerings, with less geographic coverage, variations in order turnaround and perhaps even higher prices – all of which will slow the growth curve. At least for high street retailers who have been living with seemingly endless freefall, this may be the best news in a very long time.

This article was originally published on The Conversation. Read the original article.

Why you can’t throw it all in recycling bins

We need to change the way we collect recycling to ensure that the collected items are suitable for use by manufacturers. This means changing the focus towards collecting clean, high-quality recyclable materials, segregated by type | Aspioneer
waste separation bins

For many years the recycling collected from households in the UK and other Western countries has been exported. This strategy has enabled these countries to carry on without much thought about how consumers purchase goods and dispose of all the unwanted packaging and containers. As long as there are regular collections for recycling paper, metals and plastics, little consideration is given to where this waste goes and what happens to it. But this now has to change.

Several years ago, China woke up to the environmental consequences of having the world’s recycling dumped on it to sort, process and use in manufacturing new goods. Tougher Chinese regulations came into place in 2018, aimed at improving the quality of the recycling it imported.

This should have been a wake up call to the Western world to change the way that recycling was collected and processed in order to improve the quality. But nothing changed, apart from the destination of the low-quality recycling – instead of exporting to China, the recycling was exported to several Eastern European countries and an assortment of other Asian countries, including Malaysia and the Philippines.

A dispute lasting several years over low-quality recycling that was exported by Canada to the Philippines recently saw the waste repatriated, and other countries are also set to follow this example. This solves nothing, though – this recycling has to go somewhere.

How to collect

We need to change the way we collect recycling to ensure that the collected items are suitable for use by manufacturers. This means changing the focus towards collecting clean, high-quality recyclable materials, segregated by type.

Contamination losses – which occur when non-recyclable or non-targeted materials are included in collection boxes, bags or bins and are then rejected at sorting facilities and now by overseas markets – show that keeping the end point for these materials in mind is essential.

There is often confusion over what can be recycled at home and this often varies between local authorities because they use different sorting facilities. A huge number of materials are recyclable, but the infrastructure does not exist in all parts of the UK, and sorting methods vary between these sorting facilities and this dictates which materials can be handled.

Collection systems need to be tailored towards the requirements of the reprocessors rather than the householder, collector or sorting facility – and this should guide householders on how and what to actually collect for recycling.

Efficient sorting ensures higher quality materials are collected and reaps benefits across the recycling chain. The Welsh Government Collection Blueprint, which works on this principle, secures higher quality recyclable materials with high levels of householder participation. With improved segregation of materials, clean recycling is able to enter the market for secondary materials either closer to home, or in China where it will pass tougher import requirements.

Communicating better should be a priority. Below are three ways that consumers could be helped to recycle better.

  1. Clear and consistent labelling of recyclable packaging. The removal of unhelpful “check locally” labels, which only lead to confusion and incorrect disposal of some items, is a good starting point, along with better communication of local anomalies to recycling collections.
  2. More information for householders on where recycling goes and what it is used for would improve understanding around the demand for high quality materials. The condition of recycling has an impact on its final use. If recycling is wet or greasy this often leads to contamination losses with items sticking together during mechanical sorting processes or just being so badly damaged they aren’t suitable for use.
  3. Better reporting of the end destination of recyclable materials by councils. Publishing this information is currently voluntary. This should be made mandatory and include all waste collected, not just that handled by local authorities. This would ensure consistent, ethical and legal exports. This would expose poor export practices involving low grade and poorly sorted recyclables, which helps to hide illegal exports of post-consumer electronic items which have been hidden by the lack of transparency.

Other systems

Deposit and return systems for drinks containers are being investigated by the UK government, but will soon be implemented in Scotland, where environmental issues are devolved. These systems, which operate in Norway and Germany, add a deposit to the price of drinks containers which consumers redeem when they return the empty drinks containers and can use against future purchases.

These systems collect higher quality, and there are other potential benefits, such as reducing litter, but there are also arguments against. These include the cost of implementation and the impact on existing local authority household collection schemes where the value of materials collected subsidises the cost of collecting it. But a loss of income for councils may be balanced out by a reduction in collection costs as the volume of recycling goes down.

Improving the quality of the material that is collected must be addressed, and deposit and return systems are seen as a necessary step towards doing this. Ultimately, we need a fundamental shift in the way we look at the waste we produce, which goes well beyond collecting vast quantities of it for faux recycling on the other side of the world.

This article was originally published on The Conversation. Read the original article.

Opinion: Migrants will pay the price of Mexico’s tariff deal with Trump

Approximately 80% of Mexican exports are destined for the United States. Tariffs would have devastated Mexico’s economy. To keep its goods untaxed, Mexico had to convince President Trump that it was serious about stopping migration | Aspioneer
group of people

Mexican President Andrés Manuel López Obrador is celebrating an agreement avoiding U.S. tariffs as a major political and diplomatic triumph for his government.

“We didn’t win everything, but we were able to claim a victory with there being no tariffs,” said chief negotiator Marcelo Ebrard, Mexico’s foreign affairs secretary, on June 9.

The two neighbors have been at odds since United States President Donald Trump on May 30 threatened to hit all Mexican imports with steadily rising tariffs unless Mexico successfully halted the northward flow of Central American migrants fleeing extreme poverty and violencethrough Mexico toward the United States.

Approximately 80% of Mexican exports are destined for the United States. Tariffs would have devastated Mexico’s economy.

To keep its goods untaxed, Mexico had to convince President Trump that it was serious about stopping migration. After a week of frantic negotiations, Mexico said it would deploy up to 6,000 National Guard troops to its southern border with Guatemala to stop migrants from entering Mexico.

As part of the agreement, a Trump administration program known as “Remain in Mexico,” which forces some migrants to wait in Mexico while their asylum claims are processed in the U.S., will also be expanded.

At a June 8 rally “for the dignity of Mexico and friendship with the U.S.” held in the border city of Tijuana, López Obrador pledged that Mexico will reinforce its southern border while still “applying the law and respecting the human rights” of migrants.

Ebrard added at the rally that Mexico had emerged from a near trade war with its “dignity intact.”

Doing the dirty work

As a law professor who teaches human rights, I believe that dignity will come at great cost to both Mexico and to the migrants fleeing extreme poverty and violence in Central America.

Many Mexican lawmakers, including allies of the president, have expressed outrage that their immigration policy is now bound by an “immoral and unacceptable” deal that effectively turns Mexico itself into Trump’s border wall.

The agreement violates the campaign promises of López Obrador, who took office in December promising to protect migrants’ rights and pledging not to do the U.S.‘s “dirty work” on border enforcement.

It also violates the Mexican Constitution and international law.

According to international refugee law asylum-seekers with demonstrable fear of persecution in their home countries are entitled to seek protection in the place of their choosing. Mexican law goes further. As of 2011, legitimate asylum-seekers are entitled not just to seek but to be granted asylum in Mexico.

Trump’s economic threats against Mexico may not even have been legal. Both the current North American Free Trade Agreement and the newly signed – but not yet ratified – United States-Mexico-Canada Agreement require most trade between North American countries to be tariff-free.

Even before the recent negotiations, López Obrador was already quietly complying with U.S. demands to do more to prevent migrants from reaching the U.S.

Between January and May of this year, Mexico detained 74,031 migrants – a 36% increase compared to the same period last year under former Mexican president Enrique Peña Nieto. The number of migrants deported from Mexico tripled from 5,717 in December 2018 to 15,654 in April 2019, government statistics show.

Sending troops out to target migrants, as Mexico has now promised to do, will almost certainly result in the excessive use of force against these migrants.

The Mexican National Guard is an untested new military police forcewith immigration enforcement powers. Its creation in early 2019 was highly controversial in Mexico given the Mexican military’s extraordinarily violent law enforcement record.

Between 2007 and 2017, when troops were helping police fight Mexican drug cartels, the National Human Rights Commission received over 13,700 complaints of human rights violations committed by soldiers against civilians. These included accusations of arbitrary arrests, torture and extrajudicial killings.

Mexico is not a safe country

Beyond avoiding tariffs, Mexico’s main victory in its negotiations with the United States appears to be having resisted pressure to sign a a “safe third country” agreement.

Under such an agreement, refugees are required to apply for asylum in the country where they first land and not the country where they ultimately want to settle. This means that one country can reject a person’s asylum application if they have already been granted asylum by another country.

Canada and the U.S. signed such an agreement in 2002, and Trump has been pushing Mexico to do the same since spring 2018. Under this proposal, thousands of migrants in Mexico who have already applied for asylum in the United States and are now waiting for an answer would see their applications invalidated.

Foreign Secretary Ebrard consistently rejected that proposal. He insisted that a safe third country arrangement would violate the Mexican Constitution and Mexico’s international human rights agreements.

But it is not actually clear how the newly expanded “Remain in Mexico” program – the details of which have not yet been released – will differ in practice from a safe third country agreement.

Migrants may end up staying in Mexico for years while they await their asylum hearing in the United States. During that period, Mexico will be responsible for housing, feeding and protecting refugees.

Jorge Castañeda, Mexico’s secretary of foreign affairs from 2000 to 2003, has derided Mexico’s commitment as a “light” safe third country agreement.

The mere idea that Mexico is safe is “a particularly cynical bout of wishful thinking,” as Mexican journalist León Krauze put it in a recent Washington Post op-ed.

Mexico is one of the world’s most dangerous places. An estimated 33,000 people were murdered there last year – twice the average annual homicides in the United States in a country with less than half the population.

Dozens of Central Americans have disappeared from migrant caravans journeying northward in Mexico. Over 90% of migrants say they do not feel safe in Mexico, according to a survey of 500 Central American asylum-seekers conducted in February 2018.

Mexico must do more with less

The “Remain in Mexico” policy is likely to result in a significant increase in claims filed for asylum in Mexico, where the immigration system is already under enormous strain.

Around 29,000 people applied for asylum in Mexico in 2018, according to United Nations data. This year, between January and March, Mexico received 12,716 asylum applications – 43% of last year’s total in just three months.

The Mexican Commission for Refugee Assistance, which processes asylum claims, currently has a two-year backlog.

It will now have to do more with less.

Under López Obrador’s austerity policies, the commission’s 2019 budget was cut 20%, to about US$1 million – its lowest budget since 2011.

More migrants, less money, extreme violence and a recalcitrant, unpredictable northern neighbor – these are the ingredients of a refugee crisis, not a diplomatic victory.

This article was originally published on The Conversation. Read the original article.

The problem with big data: Can’t solve social issues

As the darker sides of digital technology become clearer, public demand for more accountable, democratic, more human alternatives is growing | Aspioneer
Data cord

At almost every point in our day, we interact with digital technologies which collect our data. From the moment our smart phones wake us up, to our watch tracking our morning run, every time we use public transport, every coffee we purchase with a bank card, every song skipped or liked, until we return to bed and let our sleep apps monitor our dreaming habits – all of these technologies are collecting data.

This data is used by tech companies to develop their products and provide more services. While film and music recommendations might be useful, the same systems are also being used to decide where to build infrastructure, for facial recognition systems used by the police, or even whether you should get a job interview, or who should die in a crash with an autonomous vehicle.

Despite huge databases of personal information, tech companies rarely have enough to make properly informed decisions, and this leads to products and technologies that can enhance social biases and inequality, rather than address them.

Microsoft apologised after its chatbot started spewing hate speech“Racist” soap dispensers failed to work for people of colour. Algorithm errors caused Flickr to mislabel concentration camps as “jungle gyms”CV sorting tools rejected applications from women and there are deep concerns over police use of facial recognition tools.

These issues aren’t going unnoticed. A recent report found that 28% of UK tech workers were worried that the tech they worked on had negative consequences for society. And UK independent research organisation NESTA has suggested that as the darker sides of digital technology become clearer, “public demand for more accountable, democratic, more human alternatives is growing”.

Traditional solutions are making things worse

Most tech companies, big and small, claim they’re doing the right things to improve their data practices. Yet, it’s often the very fixes they propose that create the biggest problems. These solutions are often borne from the very same ideas, tools and technologies that got us into this mess to begin with. The master’s tools, as Audre Lorde said, will never dismantle the master’s house. Instead, we need a radically different approach from collecting more data about users, or plugging gaps with more education about digital technology.

The reason biases against women or people of colour appear in technology are complex. They’re often attributed to data sets being incomplete and the fact that the technology is often made by people who aren’t from diverse backgrounds. That’s one argument at least – and in a sense, it’s correct. Increasing the diversity of people working in the tech industry is important. Many companies are also collecting more data to make it more representative of the people who use digital technology, in the vain hope of eliminating racist soap dispensers or recruitment bots that exclude women.

The problem is that these are social, not digital, problems. Attempting to solve those problems through more data and better algorithms only serves to hide the underlying causes of inequality. Collecting more data doesn’t actually make people better represented, instead it serves to increase how much they are being surveilled by poorly regulated tech companies. The companies become instruments of classification, categorising people into different groups by gender, ethnicity and economic class, until their database looks balanced and complete.

These processes have a limiting effect on personal freedom by eroding privacy and forcing people to self-censor – hiding details of their lives that, for example, potential employers may find and disapprove of. Increasing data collection has disproportionately negative affects on the very groups that the process is supposed to help. Additional data collection leads to the over-monitoring of poorer communities by crime prediction software, or other issues such as minority neighbourhoods paying more for car insurance than white neighbourhoods with the same risk levels.

People are often lectured about how they should be careful with their personal data online. They’re also encouraged to learn how data is collected and used by the technologies that now rule their lives. While there are some merits to helping people better understand digital technologies, this approaches the problem from the wrong direction. As noted by media scholar, Siva Vaidhyanathan, this often does little more than place the burden of making sense of manipulative systems squarely onto the user, who is actually often still left powerless to do anything.

Access to education isn’t universal either. Inequalities in education and access to digital technologies means that it’s often out of reach from just those communities that are most negatively affected by social biases and the digital efforts to address them.

Social problems need social solutions

The tech industry, the media and governments have become obsessed with building ever bigger data sets to iron out social biases. But digital technology alone can never solve social issues. Collecting more data and writing “better” algorithms may seem helpful, but this only creates the illusion of progress.

Turning people’s experiences into data hides the causes of social bias – institutional racism, sexism and classism. Digital and data driven “solutions” distract us from the real issues in society, and away from examining real solutions. These digital tasks, as French philosopher Bernard Stiegler noted, only serve to increase the distance between technological systems and social organisations.

We need to slow down, stop innovating, and examine social biases not within the technology itself, but in society. Should we even build any of these technologies, or collect any of this data at all?

Better representation in the tech industry is vital, but their digital solutions will always fall short. Sociology, ethics, and philosophy have the answers to social inequality in the 21st century.

This article was originally published on The Conversation. Read the original article.

This is why electric cars may increase pollution

Evs are unlikely to help the environment, as long as they are charged using electricity generated from the same old dirty fossil fuels | Aspioneer
electric car on charging

Several countries – including France, Norway and the UK – have plans to phase out cars powered by fossil fuel before 2050, to reduce air pollution and fight climate change. The idea is to replace all conventional vehicles with electric vehicles (EVs). But this is unlikely to help the environment, as long as EVs are charged using electricity generated from the same old dirty fossil fuels.

Global electricity consumption from EVs is estimated to grow to 1,800TWh by 2040 – that’s roughly five times the current annual electricity use of UK. Using data from the UK as a benchmark, this would amount to an extra 510 megatonnes of carbon emissions coming from the electricity sector worldwide. But this massive impact could be drastically reduced if electricity is generated entirely from renewable energy sources, instead of fossil fuels.

A growing problem

To put things into perspective, 510 megatonnes is about 1.6% of the global carbon emissions in 2018. And while this may not seem like a big amount, the Intergovernmental Panel on Climate Change (IPCC) recommended that carbon emissions are reduced to net zero by 2050, to limit the average global temperature rise to 1.5°C above the pre-industrial era. So a 1.6% increase in carbon emissions is significant, and possibly catastrophic.

Perhaps this increase would be negated by the decrease in emissions, which results from phasing out polluting vehicles. But reducing global carbon emissions is not easy – in fact, emissions reached an all time high in 2018, despite the highest ever uptake of renewable energy.

Though their emissions are much lower than that of conventional cars, EVs also do generate carbon dioxide during the energy intensive manufacturing process – as do renewable energy technologies themselves.

Supply and demand

Another major issue with EVs is their impact on the availability, production and supply of rare earth metals and other scarce natural elements. EVs and their batteries contain precious metals such as lithium and cobalt. Scarcity of cobalt is already threatening the production of EVs, and alternative designs that don’t rely on scarce elements are currently being explored by car manufacturers.

This means that it’s critical to expand recycling plants dedicated to processing metals and other scarce elements for reuse. Also, detailed plans on retrofitting of conventional vehicles to turn them into EVs are needed – it’s simply not feasible to dump all conventional vehicles into landfill sites, in a scenario where they are replaced by EVs.

There are further issues with EVs that must be dealt with, if they’re to help reduce global emissions and prevent climate disaster. People are likely to charge their EVs during evening hours, after they come home from work. As more people start to use EVs, the load on the energy grid is likely to peak in the evening. And this could cause problems for electricity distribution and transmission systems, at a community or city level.

These systems may need an upgrade. Or, energy suppliers could introduce a time-of-use tariff, which is higher during peak hours and lower during off-peak times, when there’s less demand for electricity. This would encourage consumers to charge their EVs during off-peak hours.

Smart charging is another possible solution: the idea is to charge more vehicles when local electricity production through renewables such as wind and solar is high, and reduce the charging when local renewables aren’t producing enough electricity. EVs charging time can be matched with peak renewable power production using smart systems and artificial intelligence to balance the local electrical grid.

Overcoming obstacles

The high cost of EVs and the lack of available charging stations are further obstacles that the Oxford Institute for Energy Studies has identified for the mass uptake of EVs. This could create a chicken and egg scenario: the cost of EVs may not go down unless they are mass produced, and they may not be mass produced unless the costs go down. The same goes for the installation of charging stations – authorities will need foresight to recognise that extra charging stations should be built for when EV uptake increases.

Governments can help prevent these issues by subsidising EVs or providing financial incentives for clean transportation – as has already been done in China. Even on a city level, authorities can encourage people to use less polluting vehicles such as EVs through taxes or special clean air zones, as is currently being done in London.

EVs have great potential to reduce pollution and give people a more sustainable way to get around – but electricity production must also be clean. It’s not wise to rely completely on scarce natural elements required for producing EVs and alternatives have to be explored. More recycling plants are needed to make the most out of rare elements and governments need to explore ways to ensure a smooth transition to cleaner transportation.

This article was originally published on The Conversation. Read the original article.

Five tips on mentoring someone twice your age

Workplaces must prepare for an ageing workforce | Aspioneer
A woman and man sitting on a table

Plato and Aristotle. Barbara Walters and Oprah Winfrey. Steve Jobs and Mark Zuckerberg. In each of these famous relationships it was the older person with more experience acting as mentor, guiding the much younger “mentee” in their career.

But changes in the modern workplace suggest we will increasingly see more circumstances in which mentors may be younger – sometimes much younger – than their mentees.

Think back to starting a new job. Even if your workplace didn’t have a formal mentorship program – pairing you with a more experienced colleague, separate to your manager, whose role was to help you succeed – it’s likely at least one person took you “under their wing” informally.

Who will do the same for the 63-year-old returning to a workplace that looks and operates differently to the one he or she left a decade ago?

Workplaces must prepare for an ageing workforce. Twenty years ago, just a quarter of Australia’s population kept working after they turned 55. Now a third do, and the proportion will continue to rise. As people stay in the workforce longer and change jobs more often, it’s increasingly likely there will be times an older colleague might benefit from mentoring.

It isn’t even necessary to be new to an organisation. Some companies that recognise the value of staying current are embracing “reverse mentoring”, in which millennials can school older executives on technology and cultural trends.

But social norms and expectations about age and experience can make it hard for someone younger to be the mentor.

So how do you get it right?

Why it matters

Generalisations about generational differences are common. Perhaps you’ve read baby boomers (born between 1946 and 1964) value loyalty, and gen-xers (born between 1965 and 1980) work-life balance, while millennials crave innovation and change.

Such notions are more myth than fact. Stereotyping people by their membership of an age group is no less problematic than doing it according to ethnicity or gender. It can encourage unhealthy biases and create barriers to communication and understanding.

A better term than “reverse mentoring” is “inclusive mentoring”. This takes the focus off thinking there is a “natural” age order to mentoring and puts the emphasis on simply encouraging shared learning between colleagues. Everyone has something of value to learn, or teach, in a respectful environment free from age or hierarchical biases.

The key is to be conscious of the barriers. You must be aware of the stereotypes and biases – that influence expectations and perceptions to do with age, but also of the chance that different experiences can lead to different outlooks to life.

Start out by asking your colleague about their expectations of their new role, their understanding of their tasks, their past work experience, and how they anticipate the relationship going.

It’s important to remember the essentials of mentoring practice. These remain the same. A mentoring relationship is about support, sharing knowledge and insights, and being a friend. Both mentor and mentee bring something to the table.

Five top tips

  1. Understand stereotypical assumptions influence the potential success of the mentoring relationship. Tension is more likely if either of you have negative perceptions of the other based on age difference. Training in how to identify your unconscious biases might be a good idea before you start
  2. open and respectful communication should be the focus. Start the conversation by clarifying the objectives of the mentoring relationship between both of you. Being clear and focused is a good basis for mutual respect
  3. give yourselves adequate time to settle into the relationship. Your outlook towards life and work may be different. Give yourself time to get to know one another and to find common ground
  4. be open and willing to learn. You might know more about some things, but your colleague is likely to know things you don’t. Think of the mentoring relationship as a collaborative partnership where mutual learning takes place
  5. it’s OK to be apprehensive. You may feel challenged. Your colleague may feel just as uncomfortable. But with time and effort this apprehension will fade.

Becoming a good mentor, or a good mentee, isn’t automatic. It takes takes time and effort. But it is worth the effort, enriching the experience and skills of both parties, and contributing to an organisation able to compete in a changing world.

This article was originally published on The Conversation. Read the original article.

Facebook is now cleaner and group-centric, but still all about its user’s data

According to Facebook’s CEO Mark Zuckerberg, it’s “the biggest change to the app and website in the last five years” | Aspioneer
social networks

Have you noticed your Facebook feed looks different lately?

It’s a bit more “zen”, uncluttered and faster. Instagram-like story posts are displayed first, and a separate feed allows you to keep up with the latest activity in your groups.

Someone has assembled a ring of comfy chairs in your lounge room and invited the local mums and bubs group over for hot cocoa and biscuits. Even the hearts are squishier.

According to Facebook’s CEO Mark Zuckerberg, it’s “the biggest change to the app and website in the last five years”.

This cosmetic change could represent the first step in Facebook’s “privacy pivot” announced in March 2019. But we’re still waiting to hear exactly what will be happening with our data.

Pile on Facebook

Facebook has been under immense pressure from both the Federal Trade Commission in the United States, and governments around the world in the wake of a string of privacy scandals (including Cambridge Analytica).

After live-streamed terrorism in New Zealand, Jacinda Ardern is leading a global charge for regulation and oversight. The recent Christchurch Call meeting resulted in tech companies and world leaders signing an agreement to eliminate terrorist and violent extremist content online.

Everyone is piling on Facebook, even Zuckerberg’s original platform co-founder Chris Hughes.

Hughes said “it’s time to break up Facebook” and “the government must hold Mark accountable”. He was referring to the huge power Zuckerberg holds through controlling the algorithms that keep Facebook – and more recently acquired platforms Instagram and Whatsapp – ticking over. Those algorithms functionalise Facebook’s vast body of user data.

Putting it lightly

Zuckerberg admits that changes must be made, saying in April: I know we don’t exactly have the strongest reputation on privacy right now, to put it lightly.

Facebook’s business model is built on harvesting platform data about its users, crunching that to generate behavioural inferences like “divorced, male, no children, interested in weight loss”, and then selling this package to advertisers.

Technology scholar Shoshanna Zuboff calls the process of collecting and selling user data “surveillance capitalism”.

Privacy was never part of Facebook’s floor plan.

In its defence, it doesn’t sell identifiable data, and it has clamped down on developer access to its data.

That’s because developers are not the customer – nor are the users who are clicking on like buttons or buying yoga pants. Facebook’s customers are advertisers.

Facebook sells one product: a powerful capacity to personalise and target ads that is unparalleled in any other platform. This turned a profit of US$16 billion in the last quarter of 2018.

It seems reasonable to assume it’s going to do everything it can to protect its ability to keep collecting the raw material for that profit.

But recent questions put to Facebook by US Senator Josh Hawley reveal that Facebook is still not willing or able to share its plans on privacy relating to metadata collection and use.

In response to the senator, Kevin Martin, Vice President of US Public Policy at Facebook said: […] there are still many open questions about what metadata we will retain and how it may be used. We’ve committed to consult safety and privacy experts, law enforcement, and governments on the best way forward.

Chat, shop, watch … and wait

At a developer conference last month, Zuckerberg outlined his proposed changes: mainly, change the focus to communities and privacy, make messaging faster and encrypted, and transform the user experience.

The square logo is now a circle. There’s a lot of white space, and someone KonMari’d the title bar.

Shopping within Facebook is prioritised through the Marketplace feed, and you can watch shows and online videos in groups through the Watch function.

Facebook Messenger loads faster, the interface is cleaner and a dating service may soon be available in Australia.

What hasn’t changed is the core product: the capacity for Facebook to collect platform data and generate behavioural inferences for advertisers.

This article was originally published on The Conversation. Read the original article.

A self-taught AI is beating humans in a multiplayer game for the first time

This is the first time an AI has attained human-like skills in a first-person video game | Aspioneer
chess king and queen

Since the earliest days of virtual chess and solitaire, video games have been a playing field for developing artificial intelligence (AI). Each victory of machine against human has helped make algorithms smarter and more efficient. But in order to tackle real world problems – such as automating complex tasks including driving and negotiation – these algorithms must navigate more complex environments than board games, and learn teamwork. Teaching AI how to work and interact with other players to succeed had been an insurmountable task – until now.

In a new study, researchers detailed a way to train AI algorithms to reach human levels of performance in a popular 3D multiplayer game – a modified version of Quake III Arena in Capture the Flag mode.

Even though the task of this game is straightforward – two opposing teams compete to capture each other’s flags by navigating a map – winning demands complex decision-making and an ability to predict and respond to the actions of other players.

This is the first time an AI has attained human-like skills in a first-person video game. So how did the researchers do it?

The robot learning curve

In 2019, several milestones in AI research have been reached in other multiplayer strategy games. Five “bots” – players controlled by an AI – defeated a professional e-sports team in a game of DOTA 2. Professional human players were also beaten by an AI in a game of StarCraft II. In all cases, a form of reinforcement learning was applied, whereby the algorithm learns by trial and error and by interacting with its environment.

The five bots that beat humans at DOTA 2 didn’t learn from humans playing – they were trained exclusively by playing matches against clones of themselves. The improvement that allowed them to defeat professional players came from scaling existing algorithms. Due to the computer’s speed, the AI could play in a few seconds a game that takes minutes or even hours for humans to play. This allowed the researchers to train their AI with 45,000 years of gameplay within ten months of real-time.

The Capture the Flag bot from the recent study also began learning from scratch. But instead of playing against its identical clone, a cohort of 30 bots was created and trained in parallel with their own internal reward signal. Each bot within this population would then play together and learn from each other. As David Silver – one of the research scientists involved – notes, AI is beginning to “remove the constraints of human knowledge… and create knowledge itself”.

The learning speed for humans is still much faster than the most advanced deep reinforcement learning algorithms. Both OpenAI’s bots and DeepMind’s AlphaStar (the bot playing StarCraft II) devoured thousands of years’ worth of gameplay before being able to reach a human level of performance. Such training is estimated to cost several millions of dollars. Nevertheless, a self-taught AI capable of beating humans at their own game is an exciting breakthrough that could change how we see machines.

The future of humans and machines

AI is often portrayed replacing or complementing human capabilities, but rarely as a fully-fledged team member, performing the same task as human beings. As these video game experiments involve machine-human collaboration, they offer a glimpse of the future.

Human players of Capture the Flag rated the bots as more collaborative than other humans, but players of DOTA 2 had a mixed reaction to their AI teammates. Some were quite enthusiastic, saying they felt supported and that they learned from playing alongside them. Sheever, a professional DOTA 2 player, spoke about her experience teaming up with bots: “It actually felt nice; [the AI teammate] gave his life for me at some point. He tried to help me, thinking ‘I’m sure she knows what she’s doing’ and then obviously I didn’t. But, you know, he believed in me. I don’t get that a lot with [human] teammates.”

Others were less enthusiastic, but as communication is a pillar of any relationship, improving human-machine communication will be crucial in the future. Researchers have already adapted some features to make the bots more “human friendly”, such as making bots artificially wait before choosing their character during the team draft before the game, to avoid pressuring the humans.

But should AI learn from us or continue to teach themselves? Self-learning without imitating humans could teach AI more efficiency and creativity, but this could create algorithms more appropriate to tasks that don’t involve human collaboration, such as warehousing robots.

On the other hand, one might argue that having a machine trained from humans would be more intuitive – humans using such AI could understand why a machine did what it did. As AI gets smarter, we’re all in for more surprises.

This article was originally published on The Conversation. Read the original article.

Can democracy survive big tech surveillance?

The more tech companies know about their users, the more effectively they can direct them to goods and services that they are likely to buy. The more companies know about their users, the more competitive they are in the market | Aspioneer

Data is often called the oil of the 21st century.

The more tech companies know about their users, the more effectively they can direct them to goods and services that they are likely to buy. The more companies know about their users, the more competitive they are in the market.

Custom-tailored capitalism is what has made Google, Facebook, Amazon and others the richest companies in the world. This profit incentive has turned big tech into a competitive field of mass intelligence gathering. The better and more comprehensive the data, the higher profits will be.

But this business model – what I consider spying machines – has enormous potential to violate civil liberties. Big tech is already being used abroad to enhance the power of repressive regimes, as my work and others’ has shown.

While it is not presently a direct threat to U.S. democracy, I worry that the potential for future abuses exists so long as big tech remains largely unregulated.

Big tech’s spy machines

Current news is rife with examples of data abuses. In April, NBC News broke a story detailing how Facebook CEO Mark Zuckerberg had used data gathered by the platform to support his friends and defeat his rivals.

This is not Facebook’s first privacy PR nightmare. In 2018, data firm Cambridge Analytica used a Facebook app to collect data profiles of over 87 million people, which was later used to distribute targeted political advertising during elections.

Facebook is not alone in the data collection boom. This May, it was revealed that Snapchat employees were using the app’s data to obtain location data, pictures and email addresses without users’ consent. A new book by former Harvard business professor Shoshana Zuboff goes into great detail of the practices of what she calls “surveillance capitalism.” Zuboff writes, “Once we searched Google. Now, Google searches us.”

The practice goes beyond someone’s taste in music or what they purchase on Amazon. Apps created to help people through mental illness or quit smoking sell data to big tech companies. These users could be potential targets for social stigmatization or targeted advertising that exacerbates heath problems rather than solving them.

In December, The New York Times published an exposé on what one can learn about someone using their collated data from apps and smartphones. By blending location tracking with other online behavior, researchers were able to put together a detailed portrait of the most intimate details of users’ lives, such as where their children go to school or who was cheating on their diet. They could even tell which area of a nuclear power plant an individual worked in – information that is typically classified.

Because of these revelations, data that big tech collects poses a national security problem. One open source researcher used data from Strava, a fitness app, to map U.S. military bases around the world as soldiers tracked their runs. Our devices are constantly telling companies where we are and what we are doing. That is not always a good thing.

For the worst-case scenarios, look abroad

Big tech is a highly unregulated sector of the economy. Existing regulations have struggled to keep up with a rapidly innovating tech sector. In some scenarios, big tech’s capabilities are being used by dictators to craft a dystopian digital reality.

Autocratic governments around the world have already begun to use emerging technology to violate human rights. China is a prime example. China integrates AI, biometric data and online activity to track and monitor dissidents and members of ethnic minority communities, who are then sent to reeducation camps.

From my time researching the ways Russia uses these platforms to threaten democracy, I am familiar with the worst-case scenarios of big tech’s capabilities. Because platforms’ success is predicated on making information go viral, the most successful content can also be some of the most divisive. Russia believes that by disseminating enough false information about the most inflammatory areas of American politics, it can sow chaos in the system. Big tech is the perfect port of entry for such campaigns.

If Russian attacks on social media are combined with AI technology, information attacks could become precision-guided. Nefarious actors could gather the comprehensive profiles that surveillance capitalism has compiled over the years. Fake news would then no longer speak to issues but to individuals, appealing to what makes the user change their mind.

If a monopolistic tech company decided to fully embrace its capacity to spy on its users and leverage that data to a personal or political end, the consequences for democracy could be catastrophic. Americans got a taste of what an influence attack looks like during the 2016 U.S. presidential election. So long as big tech remains largely unregulated, future influence attacks on American elections will become only more potent.

Big tech isn’t going anywhere

A surface-level solution to this privacy dilemma would be for people to decouple their online lives from these companies.

For example, DuckDuckGo is an alternative search engine that does not compile user data and promises total privacy. A new browser, Brave, has promised to pay users back for selling data to advertisers.

However, these products are nowhere near as useful for a casual internet user than Google. Simply choosing not to use Google is not that simple.

While there are many different companies in question, they all hold near-monopolistic control over their corner of the market. Amazon dominates online shopping. Facebook dominates interacting with friends and causes. Google dominates web browsing.

Individuals are thus faced with a choice: Radically change their lifestyle and how they interact with the world, or continue to be the target of big tech’s spy machines.

Oversight and regulation may seem dramatic and anti-growth at the moment, but I believe that it is a necessary check on big tech – before the worst of its potentials come true.

This article was originally published on The Conversation. Read the original article.