The man who invented the World Wide Web, Tim Berners-Lee, argues that the evils of artificial intelligence have been exaggerated. ARTHUR GOLDSTUCK reports from Las Vegas.
The man who invented the World Wide Web could be mistaken for a schoolteacher, or perhaps a university professor. A slight build, spectacles and thinning brown hair combine with an almost humble demeanour that is difficult to associate with a legacy as great as any of the giants of the technology world.
Tim Berners-Lee was looking for an easier way to connect information when he first came up with the concept of the World Wide Web in the 1980s, while working as a physicist at the CERN laboratory in Switzerland. Today he is a director of the World Wide Web Consortium (W3C), which in effect sets the technical rules for how the Web operates. But he remains an academic, and is a senior researcher and founder chair at the MIT Computer Science and Artificial Intelligence Laboratory.
It is little wonder, then, that he is as preoccupied with artificial intelligence (AI) as he is with the Web. The latter remains his baby, however, and he has far more to say about it than any other topic. In March 2017, he issued an open letter warning that we have lost control of our personal data, that it’s too easy for misinformation to spread on the web, and that political advertising online needs transparency.
Clearly, he is not one to gloss over the perils of progress. So, when it was announced that he was to offer his insights into the dangers of AI at the Dell EMC 2017 conference in Las Vegas, it became the must-attend talk of an already-intensive convention.
The conference represented the first joint convention by two giants of the computer world, following computer manufacturer Dell’s purchase of storage leaders EMC for $67-billion – the biggest IT acquisition ever. EMC and its subsidiary VMware are responsible, respectively, for the storage and the management platforms of a large proportion of the world’s cloud computing infrastructure. The cloud, in turn, is going to be integral to AI, hosting and processing the massive amounts of data that will allow AI to help human beings make and act on decisions.It was no surprise that a record 13 500 delegates attended Dell EMC World. And it was no surprise that the lines to get into Berners-Lee’s “AI in Perspective” talk were almost as long as those for the conference’s opening keynote by Dell founder Michael Dell.
The Web founder did not disappoint.
While he speaks with a rapid-fire energy that sometimes appears to run ahead of his thoughts, he delivers his perspective of the future with both authority and empathy.
“The promise of AI is really exciting, but you still have to look at it as the thing which makes a lot of people concerned,” he says by way of introduction. “We have to look at not just the hopes, but also the fears.”
The promise, he says, is that of almost all computer projects:
“We’re trying to get machines to do things we don’t want to do, like filling out a form. A lot of the progress in computing starts off with simple things, like doing accounts and taxes. Translating languages has always been just about to happen, but is now starting to become functional.
“You can train a machine to beat a game. Instead of training it by looking at lots of people playing a game, you just teach it to play against it self and it becomes better than a human.
“You can grab yourself some cloud storage and some cloud computation and find some open data produced by government or scientific or enterprise, find lots of data, find the latest algorithm, and create something that has added value, and put out signal where there wasn’t signal before. Like enabling you to decide where to invest.
“More and more, computers are starting to tick off all those things we were told computers just couldn’t do. So these are very exciting times.”
There are two key problems in this Utopian vision, however. The first seems easily solved:
“There is a huge dearth of people who know how to do this stuff, but there will be more and more. The stuff is out there and you can teach yourself. The promise is huge.”
The second issue is that most difficult of challenges: public perception. Berners-Lee talks about an AI Spring, when the world was full of hope, turning into an AI winter: “The world turned on them and said: ‘You were supposed to give us robots by now, what happened? This sucks’.”
At the same time, the fear is building that AI will take away jobs.
“Suddenly its no longer AI. Now it’s Natural Language Processing, now it’s self-driving cars. But they wont call it AI.”
The three fears of AI
Berners-Lee addresses the fears of AI in “three pieces”:
“Let’s talk about robots replacing jobs. The first take you get on AI: are robots going to take my job or all the jobs of my people? Is an autonomous vehicle going to take my job? Autonomous vehicles are coming. A lot of people, when they arrive somewhere as immigrants, or people between jobs, start out with Uber or a cab company, where driving is one of the things they can do. If that goes away, there is going to be a big shift and we have to be responsible about how we do these things.”
The second big category of AI fears lies in its ability to generate fake news. However, Berners-Lee sees AI as the solution rather than the problem: “AI can be a frontline defence against things which can be proven to be false. There’s no way of really objecting to the decisions of AI.”
The third and most famous category, he says, is the Singularity – when AI surpasses human intelligence.
“Is all this getting out of control? As a kid I read Asimov books, Arthur C Clarke book. Asimov imagined robots would become just as powerful as us and therefore they’d have to be controlled. Ask people who make robots about the problem of robots becoming smarter than you, they say, ‘Do you know how difficult it is to build a robot? You know how long it’s going to take, getting smarter than us?’ Don’t worry about it.”
He mocks the current trend in movies of showing future robots not only as smart, but humanoid, female, beautiful, blonde and blue-eyed, with female voices.
“By the time they are smarter than us, they won’t look like us. A lot of the intelligence already out there is sitting in the cloud, it doesn’t have blue eyes, but it does play a part in our society.
“The funny thing is, we’re worrying about robots, but we’re not worrying about the companies that program them. People are not good at stopping bad things.”
In short, he says real people, rather than artificial intelligence, are the bigger threat.
Prepare your cam to capture the Blood Moon
On 27 July 2018, South Africans can witness a total lunar eclipse, as the earth’s shadow completely covers the moon.
Also known as a blood or red moon, a total lunar eclipse is the most dramatic of all lunar eclipses and presents an exciting photographic opportunity for any aspiring photographer or would-be astronomers.
“A lunar eclipse is a rare cosmic sight. For centuries these events have inspired wonder, interest and sometimes fear amongst observers. Of course, if you are lucky to be around when one occurs, you would want to capture it all on camera,” says Dana Eitzen, Corporate and Marketing Communications Executive at Canon South Africa.
Canon ambassador and acclaimed landscape photographer David Noton has provided his top tips to keep in mind when photographing this occasion. In South Africa, the eclipse will be visible from about 19h14 on Friday, 27 July until 01h28 on the Saturday morning. The lunar eclipse will see the light from the sun blocked by the earth as it passes in front of the moon. The moon will turn red because of an effect known as Rayleigh Scattering, where bands of green and violet light become filtered through the atmosphere.
A partial eclipse will begin at 20h24 when the moon will start to turn red. The total eclipse begins at about 21h30 when the moon is completely red. The eclipse reaches its maximum at 22h21 when the moon is closest to the centre of the shadow.
David Noton advises:
- Download the right apps to be in-the-know
The sun’s position in the sky at any given time of day varies massively with latitude and season. That is not the case with the moon as its passage through the heavens is governed by its complex elliptical orbit of the earth. That orbit results in monthly, rather than seasonal variations, as the moon moves through its lunar cycle. The result is big differences in the timing of its appearance and its trajectory through the sky. Luckily, we no longer need to rely on weight tables to consult the behaviour of the moon, we can simply download an app on to our phone. The Photographer’s Ephemeris is useful for giving moonrise and moonset times, bearings and phases; while the Photopills app gives comprehensive information on the position of the moon in our sky. Armed with these two apps, I’m planning to shoot the Blood Moon rising in Dorset, England. I’m aiming to capture the moon within the first fifteen minutes of moonrise so I can catch it low in the sky and juxtapose it against an object on the horizon line for scale – this could be as simple as a tree on a hill.
- Invest in a lens with optimal zoom
On the 27th July, one of the key challenges we’ll face is shooting the moon large in the frame so we can see every crater on the asteroid pockmarked surface. It’s a task normally reserved for astronomers with super powerful telescopes, but if you’ve got a long telephoto lens on a full frame DSLR with around 600 mm of focal length, it can be done, depending on the composition. I will be using the Canon EOS 5D Mark IV with an EF 200-400mm f/4L IS USM Ext. 1.4 x lens.
- Use a tripod to capture the intimate details
As you frame up your shot, one thing will become immediately apparent; lunar tracking is incredibly challenging as the moon moves through the sky surprisingly quickly. As you’ll be using a long lens for this shoot, it’s important to invest in a sturdy tripod to help capture the best possible image. Although it will be tempting to take the shot by hand, it’s important to remember that your subject is over 384,000km away from you and even with a high shutter speed, the slightest of movements will become exaggerated.
- Integrate the moon into your landscape
Whilst images of the moon large in the frame can be beautifully detailed, they are essentially astronomical in their appeal. Personally, I’m far more drawn to using the lunar allure as an element in my landscapes, or using the moonlight as a light source. The latter is difficult, as the amount of light the moon reflects is tiny, whilst the lunar surface is so bright by comparison. Up to now, night photography meant long, long exposures but with cameras such as the Canon EOS-1D X Mark II and the Canon EOS 5D Mark IV now capable of astonishing low light performance, a whole new nocturnal world of opportunities has been opened to photographers.
- Master the shutter speed for your subject
The most evocative and genuine use of the moon in landscape portraits results from situations when the light on the moon balances with the twilight in the surrounding sky. Such images have a subtle appeal, mood and believability. By definition, any scene incorporating a medium or wide-angle view is going to render the moon as a tiny pin prick of light, but its presence will still be felt. Our eyes naturally gravitate to it, however insignificant it may seem. Of course, the issue of shutter speed is always there; too slow an exposure and all we’ll see is an unsightly lunar streak, even with a wide-angle lens.
On a clear night, mastering the shutter speed of your camera is integral to capturing the moon – exposing at 1/250 sec @ f8 ISO 100 (depending on focal length) is what you’ll need to stop the motion from blurring and if you are to get the technique right, with the high quality of cameras such as the Canon EOS 5DS R, you might even be able to see the twelve cameras that were left up there by NASA in the 60’s!
How Africa can embrace AI
Currently, no African country is among the top 10 countries expected to benefit most from AI and automation. But, the continent has the potential to catch up with the rest of world if we act fast, says ZOAIB HOOSEN, Microsoft Managing Director.
To play catch up, we must take advantage of our best and most powerful resource – our human capital. According to a report by the World Economic Forum (WEF), more than 60 percent of the population in sub-Saharan Africa is under the age of 25.
These are the people who are poised to create a future where humans and AI can work together for the good of society. In fact, the most recent WEF Global Shapers survey found that almost 80 percent of youth believe technology like AI is creating jobs rather than destroying them.
Staying ahead of the trends to stay employed
AI developments are expected to impact existing jobs, as AI can replicate certain activities at greater speed and scale. In some areas, AI could learn faster than humans, if not yet as deeply.
According to Gartner, while AI will improve the productivity of many jobs and create millions more new positions, it could impact many others. The simpler and less creative the job, the earlier, a bot for example, could replace it.
It’s important to stay ahead of the trends and find opportunities to expand our knowledge and skills while learning how to work more closely and symbiotically with technology.
Another global study by Accenture, found that the adoption of AI will create several new job categories requiring important and yet surprising skills. These include trainers, who are tasked with teaching AI systems how to perform; explainers, who bridge the gap between technologist and business leader; and sustainers, who ensure that AI systems are operating as designed.
It’s clear that successfully integrating human intelligence with AI, so they co-exist in a two-way learning relationship, will become more critical than ever.
Combining STEM with the arts
Young people have a leg up on those already in the working world because they can easily develop the necessary skills for these new roles. It’s therefore essential that our education system constantly evolves to equip youth with the right skills and way of thinking to be successful in jobs that may not even exist yet.
As the division of tasks between man and machine changes, we must re-evaluate the type of knowledge and skills imparted to future generations.
For example, technical skills will be required to design and implement AI systems, but interpersonal skills, creativity and emotional intelligence will also become crucial in giving humans an advantage over machines.
“At one level, AI will require that even more people specialise in digital skills and data science. But skilling-up for an AI-powered world involves more than science, technology, engineering and math. As computers behave more like humans, the social sciences and humanities will become even more important. Languages, art, history, economics, ethics, philosophy, psychology and human development courses can teach critical, philosophical and ethics-based skills that will be instrumental in the development and management of AI solutions.” This is according to Microsoft president, Brad Smith, and EVP of AI and research, Harry Shum, who recently authored the book “The Future Computed”, which primarily deals with AI and its role in society.
Interestingly, institutions like Stanford University are already implementing this forward-thinking approach. The university offers a programme called CS+X, which integrates its computer science degree with humanities degrees, resulting in a Bachelor of Arts and Science qualification.
Revisiting laws and regulation
For this type of evolution to happen, the onus is on policy makers to revisit current laws and even bring in new regulations. Policy makers need to identify the groups most at risk of losing their jobs and create strategies to reintegrate them into the economy.
Simultaneously, though AI could be hugely beneficial in areas such as curbing poor access to healthcare and improving diagnoses for example, physicians may avoid using this technology for fear of malpractice. To avoid this, we need regulation that closes the gap between the pace of technological change and that of regulatory response. It will also become essential to develop a code of ethics for this new ecosystem.
Preparing for the future
With the recent convergence of a transformative set of technologies, economies are entering a period in which AI has the potential overcome physical limitations and open up new sources of value and growth.
To avoid missing out on this opportunity, policy makers and business leaders must prepare for, and work toward, a future with AI. We must do so not with the idea that AI is simply another productivity enhancer. Rather, we must see AI as the tool that can transform our thinking about how growth is created.
It comes down to a choice of our people and economies being part of the technological disruption, or being left behind.