recentpopularlog in

aries1988 : ai   96

« earlier  
DeepMind and Google: the battle to control artificial intelligence
Hassabis thought DeepMind would be a hybrid: it would have the drive of a startup, the brains of the greatest universities, and the deep pockets of one of the world’s most valuable companies. Every element was in place to hasten the arrival of AGI and solve the causes of human misery.

Demis Hassabis was born in north London in 1976 to a Greek-Cypriot father and a Chinese-Singaporean mother. He was the eldest of three siblings. His mother worked at John Lewis, a British department store, and his father ran a toy shop. He took up chess at the age of four, after watching his father and uncle play. Within weeks he was beating the grown-ups. By 13 he was the second-best chess player in the world for his age. At eight, he taught himself to code on a basic computer.

Hassabis officially founded DeepMind on November 15th 2010. The company’s mission statement was the same then as it is now: to “solve intelligence”, and then use it to solve everything else. As Hassabis told the Singularity Summit attendees, this means translating our understanding of how the brain accomplished tasks into software that could use the same methods to teach itself.

It’s an impressive demo. But Hassabis leaves a few things out. If the virtual paddle were moved even fractionally higher, the program would fail. The skill learned by DeepMind’s program is so restricted that it cannot react even to tiny changes to the environment that a person would take in their stride – at least not without thousands more rounds of reinforcement learning. But the world has jitter like this built into it. For diagnostic intelligence, no two bodily organs are ever the same. For mechanical intelligence, no two engines can be tuned in the same way. So releasing programs perfected in virtual space into the wild is fraught with difficulty.
google  ai  story  uk  game  startup  reportage 
march 2019 by aries1988
GE Says It’s Leveraging Artificial Intelligence To Cut Product Design Times In Half

It’s a neural network that’s trained with the results of standard two-day computational fluid dynamics (CFD) analyses of variations in a particular design to estimate the conclusions that a CFD would come to. In one test case, in which the researchers trained the surrogate model with about 100 CFDs to figure out the optimum shape for the crown of a piston in a diesel engine, the model was able to evaluate roughly a million design variations in 15 minutes, an increase in speed of 5 billion times. More typically the researchers expect to achieve an improvement of 10 million to 100 million times. The best design of the piston crown produced a 7% improvement in fuel efficiency with a “significant” reduction in soot emissions, they say.

“We can, say, take all the knowledge that went into designing the GE9X or the LEAP [jet engines] and apply it to developing a hypersonic or apply it to a next-gen narrow-body,” says Tallman. “We’re confident that it will provide insights that we wouldn’t have otherwise.”
industry  design  ai 
march 2019 by aries1988
Homo sapiens devient-il homo informaticus ?
oisif, oisive

adjectif et nom
(ancien français oidif, avec l'influence de oiseux)

Qui n'exerce aucune activité permanente et dispose de nombreux loisirs, qui vit sans travailler.
human  workforce  future  book  français  opinion  work  self  ai  data 
september 2018 by aries1988
Yuval Noah Harari on Why Technology Favors Tyranny - The Atlantic

- In 2018 the common person feels increasingly irrelevant.
By 2050, a useless class might emerge, the result not only of a shortage of jobs or a lack of relevant education but also of insufficient mental stamina to continue learning new skills.

- whatever liberal democracy’s philosophical appeal, it has gained strength in no small part thanks to a practical advantage: The decentralized approach to decision making that is characteristic of liberalism—in both politics and economics.
In the late 20th century, democracies usually outperformed dictatorships, because they were far better at processing information.
Democracy distributes the power to process information and make decisions among many people and institutions, whereas dictatorship concentrates information and power in one place.
- If you disregard all privacy concerns and concentrate all the information relating to a billion people in one database, you’ll wind up with much better algorithms than if you respect individual privacy and have in your database only partial information on a million people.

- What will happen to this view of life as we rely on AI to make ever more decisions for us?
once we begin to count on AI to decide what to study, where to work, and whom to date or even marry, human life will cease to be a drama of decision making, and our conception of life will need to change. Democratic elections and free markets might cease to make sense. So might most religions and works of art.
If we are not careful, we will end up with downgraded humans misusing upgraded computers to wreak havoc on themselves and on the world.

- For starters, we need to place a much higher priority on understanding how the human mind works—particularly how our own wisdom and compassion can be cultivated.
- More practically, and more immediately, if we want to prevent the concentration of all wealth and power in the hands of a small elite, we must regulate the ownership of data.
advice  future  crisis  ai  society  politics  people  life  work  mentality  human  democracy  dictatorship  competition  liberalism 
september 2018 by aries1988
Is the Algorithmification of the Human Experience a Good Thing?
Skeptics will point out that those algorithms are designed by corporations to serve their interests, not yours. Social media companies, for instance, want to keep you on their services as long as possible, which makes them prone to pushing emotionally charged content that might not be super healthy for you or for society. And even a benevolent algorithm can produce negative or unwanted results.

The video wasn’t directly crafted by a machine. But it wasn’t totally a human creation, either. Rather, it was optimized to appeal to YouTube’s content algorithm, which automatically plays related videos one after the other. “Johny Johny Yes Papa” copies enough elements of popular kid’s videos — a certain length, musical beat, color palette and visual style, along with key words and lyrics — that YouTube’s algorithm will line it up after more popular songs.

At least on a symbolic level, there is something unsettling about a global, faceless content empire that hoovers up human culture and processes it into homogenized nothingness to be fed to kids via tireless social media algorithms that seek, above all else, to maximize time spent on site.
society  engineering  crisis  question  future  children  video  ai  algorithm 
september 2018 by aries1988
Machine Learning
Reading list
Kevin P. Murphy. Machine Learning: A Probabilistic Perspective. MIT Press 2012
Ian Goodfellow, Yoshua Bengio and Aaron Courville. Deep Learning. MIT Press 2016
primer  ai  course 
july 2018 by aries1988
由劍橋生化博士到初創企業家,她要用AI推翻科學知識的高牆 |端傳媒 Initium Media
entrepreneurial  ai  research  academia 
july 2018 by aries1988
What if the Government Gave Everyone a Paycheck? - The New York Times
A world inhabited only by robots, their billionaire owners and a large and increasingly restive population is the plotline for countless dystopian fantasies, but it’s a reality that appears to be drawing closer. If we continue on the path we’re on, we will need to make fundamental choices about how to support human livelihoods and ensure equal participation in our economy and society. Most basically, we will have to confront the realities of vastly unequal economic and political power. Even if we manage to enact a U.B.I., it will not be nearly enough.
poverty  ai  future  money  wealth  state 
july 2018 by aries1988
在 D 版发过了,不过因为不少朋友看不到 D 版,我就放在这里吧,说说我最近做的这个 Project - V2EX
分享创造 - @coolwulf - 去年的时候,我一个在芝加哥比我小几级的南京大学校友去世了。乳腺癌,发现得晚了,才 34 岁,留下了一个 4 岁的孩子。非常可惜。想想能不能做点什么事情可以帮助大众来提高乳腺癌的早期检测成功率。因为如果
ai  health  online  cancer  tool 
june 2018 by aries1988
RT : 'Labor shortage and low productivity were threatening the future of the Kato farm, on the Japanese island of Hokkai…
ai  video  japan  farming  population  crisis  youth  workforce  research  reportage  fromage  milk  robot  children  family 
may 2018 by aries1988
BBC World Service - Business Daily, TED 2018: Changing the AI Conversation
Do we really know the potential and the pitfalls of artificial intelligence?
TED  2018  ai  podcast 
april 2018 by aries1988
4 月 17 日和 18 日,美国著名学者、作家侯世达(Douglas Richard Hofstadter)将来中国参与两场活动,并宣传他即将出版的中文新书《表象与本质:类比作为思维的燃料与火焰》(Surfaces and Esse...
book  ai 
april 2018 by aries1988
We Know How You Feel
just as the increasing scarcity of oil has led to more exotic methods of recovery, the scarcity of attention, combined with a growing economy built around its exchange, has prompted R. & D. in the mining of consumer cognition. “What people in the industry are saying is ‘I need to get people’s attention in a shorter period of time,’ so they are trying to focus on capturing the intensity of it,” Teixeira explained. “People who are emotional are much more engaged. And because emotions are ‘memory markers’ they remember more. So the idea now is shifting to: how do we get people who are feeling these emotions?”
emotion  business  research  story  egypt  female  scientist  startup  robot  ai  future 
february 2018 by aries1988
Self-driving cars face a new test: snow
Autonomous vehicles must overcome technical difficulties to manage winter weather
snow  winter  ai  car  driving 
december 2017 by aries1988
La traduction dopée par l’intelligence artificielle
« Tout le monde s’est rué sur ces technologies. C’était complètement fou ! », raconte Philipp Koehn, de l’université Johns-Hopkins (Maryland), pionnier d’une technique précédente, balayée par la nouvelle venue. « Avant ces inventions, on estimait qu’il fallait un an pour progresser d’un point sur une certaine échelle de qualité. Après, en un an, les bonds, pour certaines paires de langues, ont été de près de huit points », constate François Yvon, ­directeur du Laboratoire d’informatique pour la mécanique et les sciences de l’ingénieur (Limsi-CNRS) à Orsay (Essonne). Et en août, un nouveau venu, DeepL, aussi à l’origine du dictionnaire Linguee, se targuait d’un gain de trois points supplémentaires sur la même échelle de qualité par rapport à ses concurrents.

Puis nouvel hiver dans le domaine, avec des évolutions assez lentes. Jusqu’aux secousses de l’année 2014. Trois articles, quasi simultanés, l’un de chercheurs de Google, les deux autres de l’équipe de l’université de Montréal menée par Yoshua Bengio, expliquent comment de nouveaux algorithmes promettent de tout changer. Les mots-clés ne sont plus « linguistique » ou « statistique » mais « apprentissage » et « réseaux de neurones ». Ces derniers ont été inventés dans les années 1950 et remis au goût du jour, notamment par Yoshua Bengio, pour la reconnaissance de caractères manuscrits ou l’identification ­d’objets ou d’animaux dans les images.

« Formellement, apprendre, pour ces réseaux, c’est évaluer les paramètres de cette fonction qui associe une phrase source à une phrase cible », ­résume François Yvon.
reportage  ai  translation  literature  communication  history  today 
november 2017 by aries1988
Dawn of the New Everything by Jaron Lanier — virtual virtues
The most optimistic argument in this book is that people will still produce the data on which AI systems are trained. That makes the algorithms only an imperfect derivative of our human world, and by definition inferior. They will only be credible to the extent that we choose to bow to them.
VR  book  ai 
november 2017 by aries1988
“人工智能”一剂强心针,能让埋头写考卷的中国学生更加聪明吗?|深度|端传媒 Initium Media


ai  hefei  school 
november 2017 by aries1988
Why machines do not have to be the enemy
There are risks to workers from smarter computers, but human skills still have value
ai  numbers 
november 2017 by aries1988
La « nouvelle ère » Xi, un défi pour le modèle occidental
Enfin, seul maître à bord, Xi promet à son 1,4 milliard de compatriotes un « développement en deux étapes » : d’ici à 2035, la Chine aura achevé sa modernisation, notamment en termes d’innovation et, en 2049, année du centenaire de la fondation de la République populaire, elle aura atteint le statut de leader planétaire, défendue par une armée « de premier rang mondial ».

Il lui manque toujours la reconnaissance d’un prix Nobel, les scandales de fraude sont encore trop nombreux et l’argent ne peut pas tout. Mais la taille compte. Avec 730 millions de personnes connectées, un usage du téléphone mobile plus avancé que celui des pays occidentaux et infiniment moins de barrières éthiques, la Chine aborde la bataille de l’intelligence artificielle avec de gros atouts.

Si « l’ère » est nouvelle, cependant, le modèle de la concentration des pouvoirs dans les mains d’un seul homme et de son parti, lui, est familier. Cela s’appelle une dictature. Son succès serait, pour le coup, une authentique innovation.
chronique  china  2017  future  politics  innovation  technology  comparison  ai 
october 2017 by aries1988
The AI That Has Nothing to Learn From Humans - The Atlantic
From all accounts, one gets the sense that an alien civilization has dropped a cryptic guidebook in our midst: a manual that’s brilliant—or at least, the parts of it we can understand.

You can see Go as a massive tree made of thousands of branches representing possible moves and countermoves. Over generations, Go players have identified certain clusters of branches that seem to work really well. And now that AlphaGo’s come along, it’s finding even better options. Still, huge swaths of the tree might yet be unexplored. As Lockhart put it, “It could be possible that a perfect God plays [AlphaGo] and crushes it. Or maybe not. Maybe it’s already there. We don't know.”

“Generally the way humans learn Go is that we have a story,” he points out. “That’s the way we communicate. It’s a very human thing.”

After all, people can identify and discuss shapes and patterns. Or we can argue with each other about the reasons a killer move won the game. Take a basic example: When teaching beginners, a Go instructor might point out an odd-looking formation of stones resembling a lion’s mouth or a tortoiseshell (among other patterns) and discuss how best to play in these situations. In theory, AlphaGo could have something akin to that knowledge: A portion of its neural network might hypothetically be “sounding an alarm,” so to speak, whenever that lion’s-mouth pattern appears on the board. But even if that were the case, AlphaGo isn’t equipped to turn this sort of knowledge into any kind of a shareable story. So far, that task is one that still falls to people.
ai  human  comparison  game  go  2017 
october 2017 by aries1988
The West should stop worrying about China’s AI revolution
China has some big advantages in AI. It has a wealth of talented engineers and scientists, for one. It also is rich in the data necessary to train AI systems. With fewer obstacles to data collection and use, China is amassing huge databases that don’t exist in other countries. The results can be seen in the growth of facial-recognition systems based on machine learning: they now identify workers at offices and customers in stores, and they authenticate users of mobile apps.

The location of the institute is well chosen. From the office windows, you can see the campuses of both Peking University and Tsinghua University, two of China’s top academic institutions. Sinovation provides machine-learning tools and data sets to train Chinese engineers, and it offers expertise for companies hoping to make use of AI. The institute has about 30 full-time employees so far, but the plan is to employ more than 100 by next year, and to train hundreds of AI experts each year through internships and boot camps. Right now, roughly 80 percent of the institute’s funding and projects are aimed at commercializing AI, while the rest is focused on more far-out technology research and startup incubation.

The goal isn’t to invent the next AlphaGo, though; it’s to upgrade thousands of companies across China using AI.

“The titans of industry [in China] have seen fortunes made and fortunes lost all within their lifetime,” he says. “When you see the tech trends shift, you had better move quickly, or someone else will beat you.”
ai  china  list  entrepreneurial  company  internet  innovation  today  comparison  opinion  reportage  shenzhen 
october 2017 by aries1988
LinkedIn founder Reid Hoffman: ‘Board games inspired my business strategy’

Settlers of Catan is part of a group of so-called “German-style board games” which reward strategy rather than luck and are less centred on themes of conflict than many US board games. Devised in 1995 by designer Klaus Teuber, it has also been reimagined as a very popular app. Set on a fictional island in Viking times, the aim is to collect and trade commodity cards (such as wool, grain and brick), before exchanging them for plastic roads and settlements to occupy the board. Points are awarded for things like having the longest road, and the first player to reach 10 points wins.

He says he prefers games to that other great standby of American males, hanging out watching sports. “People are bad about social stuff. They get uncomfortable in silence. One of the benefits of a board game is it replaces the silence, it keeps the momentum of the conversation going.”

Discussing books he has read recently, he enthuses about Nonzero by Robert Wright — “one of my favourite intellectual authors. Basically, his theory is you have cultural evolution because you have a preference for non-zero sum games.” As society evolves, there are more and more interactions where both sides come out a winner.
game  comparison  technology  siliconvalley  american  entrepreneurial  politics  human  ai  thinking  future 
october 2017 by aries1988
The Seven Deadly Sins of AI Predictions - MIT Technology Review

We see a similar pattern with other technologies over the last 30 years. A big promise up front, disappointment, and then slowly growing confidence in results that exceed the original expectations. This is true of computation, genome sequencing, solar power, wind power, and even home delivery of groceries.

modern-day AGI research is not doing well at all on either being general or supporting an independent entity with an ongoing existence. It mostly seems stuck on the same issues in reasoning and common sense that AI has had problems with for at least 50 years. All the evidence that I see says we have no real idea yet how to build one. Its properties are completely unknown, so rhetorically it quickly becomes magical, powerful without limit.

Today’s machine learning is not at all the sponge-like learning that humans engage in, making rapid progress in a new domain without having to be surgically altered or purpose-built.

What Gordon Moore actually said was that the number of components that could fit on a microchip would double every year. That held true for 50 years, although the time constant for doubling gradually lengthened from one year to over two years, and the pattern is coming to an end.

The U.S. Air Force still flies the B-52H variant of the B-52 bomber. This version was introduced in 1961, making it 56 years old. The last one was built in 1962, a mere 55 years ago. Currently these planes are expected to keep flying until at least 2040, and perhaps longer—there is talk of extending their life to 100 years.
robot  ai  future  technology  opinion 
october 2017 by aries1988
Aerodynamics For Cognition |
By studying how birds fly and the structure of their wings, you can learn something important about aerodynamics. And what you learn about aerodynamics is equally relevant to then being able to make jet engines.                                 

The kind of work that I do is focused on trying to identify the equivalent of aerodynamics for cognition. What are the real abstract mathematical principles that constrain intelligence? What can we learn about those principles by studying human beings? 

We already do this to some extent. If you’ve ever used the strategy of gamification, where you’re using an app or something that gives you points for completing tasks, or if you make a to-do list and you get satisfaction from checking things off, what you’re doing is essentially using this external device as a mechanism for changing the environment that you’re in.

What machine-learning algorithms do when they're solving this problem is recognize that the thing you should be doing is exploring more when you first arrive in the city and exploiting more the longer you are in the city. The value of that new information decreases over time. You're less likely to find a place that is better than the places you've seen so far, and the number of opportunities that you're going to have to exploit that knowledge is decreasing.

My colleague Alison Gopnik, who has been pursuing this, has a hypothesis about cognitive development. When we look at children, that variability and randomness that we see is exactly a rational response to the structure of the problems they're trying to solve. If they're trying to figure out what are the things in their environment that they will most enjoy, then putting everything in their mouth is a pretty good strategy in terms of maximizing their exploration.

In the first half of the 20th century, it was disreputable to try to study how the mind works because minds were things that you never saw or touched or intervened on. What you could see was behavior and the environment that induces that behavior, so the behaviorist psychologists said, "Let's get rid of the mind. Let's just focus on these mappings from environment to behavior." That's where a lot of behavioral data science is. If I show you this, then you click on this. If you've seen these webpages, then you're likely to go to this webpage. It's a very behaviorist conception of what underlies the way that people are acting.

In Australia, in the last year of high school, you have to make a decision about what you want to study at university. It was 1994, I was sixteen years old, and I had no idea what I wanted to do. I knew that I liked math, but I certainly didn't want to make a commitment to doing that for the rest of my life. I said, "Okay, I'll study the things that we don't know anything about—philosophy, psychology, anthropology." That was what I went to university to do.

One of the ways in which human beings still outperform computers is in being able to solve problems of reasoning about why you did the thing you did, what you're going to do next, what the underlying reasons were behind things that you did.

We as human beings are used to being surrounded by intelligent systems whose thoughts are opaque to us. It's just that normally those intelligent systems are human beings.
ai  thinking  research  human  interaction  communication  motivation  consciousness  brain  maths 
october 2017 by aries1988
人工智能专家 Max Tegmark:机器将会超越人类,但这未必是一件坏事 | 爱范儿
问题在于,如果超级智能真的出现,人类将会如何看待自身?超级智能将会定义我们的生活,而人类也不再是改变世界的主要力量?“是这样的,”Tegmark 说,“但是,世界上的许多人已经有类似的想法,并且为此感到幸福。信教的人相信,更强大、更聪明的神明关照着他们。我觉得,我们应该放弃一种高傲的想法,即自我价值应该建立在人类例外论上。那是一种错误观点。如果我们变得谦卑一点,承认自己并非最聪明的生物,那么,我们的生活将会变得更好。我们可以从其他东西里获得价值:与其他人建立更加深刻的联系,去体验一些更加美好的事物。”
october 2017 by aries1988
Superintelligence: a space odyssey
According to physicist Max Tegmark, ‘Life 3.0’ is set to expand into the cosmos. Will there be a place for humanity?

Life 3.0 is essentially the story of his intellectual journey over the following three years, as he moves from fatalistic gloom to a new optimism. Tegmark remains convinced that barring some cataclysmic disaster in the next few decades, superhuman AI will take over the world. But he now believes that we can shape the way this happens, in a way that embodies positive human values.

Once AI has exceeded human abilities, our knowledge of physics suggests that it will advance rapidly beyond the point that biological intelligence has reached through random evolutionary progress. As Tegmark points out, “information can take on a life of its own, independent of its physical substrate”. In other words any aspect of intelligence — presumably including consciousness — that evolved in flesh, blood and carbon atoms can exist in silicon or any other material.

The fundamental limit imposed by the laws of physics is a billion trillion trillion times more powerful than today’s best computers.
ai  book 
october 2017 by aries1988
Yuval Noah Harari : « La technologie nous laisse le choix, à condition d’être imaginatifs »
je propose une vision globale des phénomènes. Les gens sont submergés par les informations nouvelles. Ils n’en veulent donc pas davantage, mais souhaitent que quelqu’un les structure. Je suis un peu comme Google et son moteur de recherche qui organisent la Toile !

Les dictatures à venir, nourries par une masse de données, n’oppresseront plus ces groupes mais les individus eux-mêmes, dont on saura tout. Il sera plus difficile de résister à des discriminations pour l’accès au logement, au crédit, à l’emploi, car on sera seul et non plus membre d’un groupe maltraité. En plus, l’algorithme aura sans doute raison ! On est piégé. Bref, nous devrons affronter des crises bien avant l’avènement d’une superintelligence qui remplacerait les hommes.

Quelle que soit la réponse, ce n’est pas très important : ce qui compte, c’est que des gens y croient. Ce ne serait pas la première fois que des idées fausses mènent le monde. Au XXe siècle, le darwinisme social a eu des effets politiques et sociaux très importants, alors que des scientifiques savaient que cette pensée était fondée sur une conception erronée du darwinisme en biologie. De même, toutes les religions proposent une vue déformée de la réalité, mais elles convainquent les gens et ont changé le monde.
book  buy  interview  ai  future  human 
september 2017 by aries1988
Superintelligence: a space odyssey

No one knows what the next blockbuster substrate will be but Tegmark is confident that the doubling of computing power every couple of years will continue indefinitely. The fundamental limit imposed by the laws of physics is a billion trillion trillion times more powerful than today’s best computers.

The message at the heart of Life 3.0 and Tegmark’s beneficial AI movement is that, since super-AI is almost inevitable, we should make every effort now to ensure that it emerges in a way that will be as friendly as possible to human beings — primed to deliver the cosmic inheritance we want. If we wait too long, we may be too late.

At present no one has a clear idea how to achieve this. On the moral and political level we need to discuss what goals and qualities to incorporate. On the technical and scientific level, researchers must figure out how to build our chosen human values into AI in a way that will preserve them after we have lost direct control of its development. Tegmark advances various options and scenarios in which superintelligence plays roles ranging from gatekeeper to protector god, zookeeper to enslaved god.
ai  book  opinion  future  human 
august 2017 by aries1988
The Future of Intelligence
In this episode of the Waking Up podcast, Sam Harris speaks with Max Tegmark about his new book Life 3.0: Being Human in the Age of Artificial Intelligence. They talk about the nature of intelligence, the risks of superhuman AI, a…

Max Tegmark is a professor of physics at MIT and the co-founder of the Future of Life Institute. Tegmark has been featured in dozens of science documentaries. He is the author of Our Mathematical Universe and Life 3.0: Being Human in the Age of Artificial Intelligence.
intelligence  ai  human 
august 2017 by aries1988
D.I.Y. Artificial Intelligence Comes to a Japanese Family Farm
The Koikes have been growing cucumbers in Kosai, a town wedged between the Pacific Ocean and the brackish Lake Hamana, for nearly fifty years.

For his project, he used TensorFlow, which Google released to the public in 2015.

He began by building a custom photo stand, which allowed him to photograph each cucumber from three angles. Then, to analyze the images, he adapted a popular piece of TensorFlow software used for recognizing handwritten numerals. Before he could turn the A.I. loose, though, Koike had to train it. He captured seven thousand photos of cucumbers that his mother had already sorted, then used the data to teach his software to recognize which vegetables belonged in which categories. Finally, he built an automated conveyor-belt system to move each cucumber from the photo stand to the bin designated by the program.
agriculture  ai  business  city  countryside  engineering  example  family  japanese  story 
august 2017 by aries1988
How Checkers Was Solved
“From the end of the Tinsley saga in ’94–’95 until 2007, I worked obsessively on building a perfect checkers program,” Schaeffer told me. “The reason was simple: I wanted to get rid of the ghost of Marion Tinsley. People said to me, ‘You could never have beaten Tinsley because he was perfect.’ Well, yes, we would have beaten Tinsley because he was only almost perfect. But my computer program is perfect.”

And then there is his most quotable line: “Chess is like looking out over a vast open ocean; checkers is like looking into a bottomless well.”
ai  competition  duel  engineering  game  genius  human  maths  story 
august 2017 by aries1988
Practical Deep Learning For Coders—18 hours of lessons for free's practical deep learning MOOC for coders. Learn CNNs, RNNs, computer vision, NLP, recommendation systems, keras, theano, and much more! neural networks!
ai  learn  self  python 
may 2017 by aries1988
BBC World Service - Business Daily, Machine Learning

Machines are about to get a lot smarter and machine learning will transform our lives. So says a report by the Royal Society in the UK, a fellowship of many of the world’s most eminent scientists. Machine learning is a form of artificial intelligence that’s already being used to tag people in photos, to interpret voice commands and to help internet retailers to make recommendations.
ai  interview  expert  explained  challenge  today 
april 2017 by aries1988
There’s a big problem with AI: even its creators can’t explain how it works

In 2015, researchers at Google modified a deep-learning-based image recognition algorithm so that instead of spotting objects in photos, it would generate or modify them. By effectively running the algorithm in reverse, they could discover the features the program uses to recognize, say, a bird or building. The resulting images, produced by a project known as Deep Dream, showed grotesque, alien-like animals emerging from clouds and plants, and hallucinatory pagodas blooming across forests and mountain ranges.
ai  health  medical  cancer  problem  communication  today  human  google  art  visualization  algorithm 
april 2017 by aries1988
The mind in the machine: Demis Hassabis on artificial intelligence

Top human Go players deal with this enormous complexity by leaning heavily on their intuition and instinct, often describing moves as simply feeling right, in contrast to chess players, who rely more on precise calculation.

the consensus among experts that this breakthrough was a decade ahead of its time. More importantly, during the games AlphaGo played a handful of highly inventive winning moves, one of which — move 37 in game two — was so surprising it overturned hundreds of years of received wisdom and has been intensively examined by players since. In the course of winning, AlphaGo somehow taught the world completely new knowledge about perhaps the most studied game in history.

We must also continue to be highly cognisant of both the utility and limitations of AI algorithms.
april 2017 by aries1988
Westworld OST - Reveries (by Ramin Djawadi) - YouTube
At the end of the day in Tokyo, try putting this on repeat and falling asleep. Surreal.
Westworld OST - Reveries
tv  music  ai 
march 2017 by aries1988
Google capable de « dépixelliser » des images
Des chercheurs en intelligence artificielle ont développé une technologie permettant de recomposer un visage. Mais pas au point d’identifier une personne.
ai  image  photo 
february 2017 by aries1988
How robots are making humans indispensable

Mr Shestakovsky initially assumed that his research would show how machines were replacing human workers. When he did grassroots analysis he realised that the company was growing so fast, with such big and complex computing systems, that it was constantly drafting more humans — not robots — to monitor, manage and interpret the data. Software automation can substitute for labour but it also creates new human-machine complementaries, he told an American Anthropological Association meeting recently, noting that companies are creating new types of jobs.

new digitised jobs may seem less attractive than the old roles since they are often structured as contingent work, with self-employed workers who provide services on demand.

the urgent need for a bigger policy debate about how to prepare workers for this new world. Workforce training needs to change to instil more digital skills. New types of social security, health and pension systems are necessary to accommodate contingent workers.
automation  workflow  usa  robot  ai 
december 2016 by aries1988
Our Automated Future

"I for one welcome our new computer overlords."

Could another person learn to do your job by studying a detailed record of everything you’ve done in the past? Martin Ford, a software developer, asks early on in Rise of the Robots: Technology and the Threat of a Jobless Future (Basic Books).

Imagine a matrix with two axes, manual versus cognitive and routine versus nonroutine. Jobs can then be arranged into four boxes: manual routine, manual nonroutine, and so on. (Two of Brynjolfsson and McAfee’s colleagues at M.I.T., Daron Acemoglu and David Autor, performed a formal version of this analysis in 2010.) Jobs on an assembly line fall into the manual-routine box, jobs in home health care into the manual-nonroutine box. Keeping track of inventory is in the cognitive-routine box; dreaming up an ad campaign is cognitive nonroutine.

Later, Ford notes, A computer doesn’t need to replicate the entire spectrum of your intellectual capability in order to displace you from your job; it only needs to do the specific things you are paid to do.

Each new technology displaced a new cast of workers: first knitters, then farmers, then machinists. The world as we know it today is a product of these successive waves of displacement, and of the social and artistic movements they inspired: Romanticism, socialism, progressivism, Communism.

Even as robots grow cleverer, some tasks continue to bewilder them. At present, machines are not very good at walking up stairs, picking up a paper clip from the floor, or reading the emotional cues of a frustrated customer

Routine jobs on the factory floor or in payroll or accounting departments tend to fall in between. And it’s these middle-class jobs that robots have the easiest time laying their grippers on.

As recently as twenty years ago, Google didn’t exist, and as recently as thirty years ago it couldn’t have existed, since the Web didn’t exist. At the close of the third quarter of 2016, Google was valued at almost five hundred and fifty billion dollars and ranked as the world’s second-largest publicly traded company, by market capitalization. (The first was Apple.)

Google also illustrates how, in the age of automation, new wealth can be created without creating new jobs. Google employs about sixty thousand workers. General Motors, which has a tenth of the market capitalization, employs two hundred and fifteen thousand people. And this is G.M. post-Watson. In the late nineteen-seventies, the carmaker’s workforce numbered more than eight hundred thousand.
ai  automation  robot  manufacturing  industry  workforce  future  crisis  opportunity  politics  book  opinion  prediction  history  explained  watson 
december 2016 by aries1988
The Great A.I. Awakening -

differentiates between the current applications of A.I. and the ultimate goal of artificial general intelligence. Artificial general intelligence will not involve dutiful adherence to explicit instructions, but instead will demonstrate a facility with the implicit, the interpretive. It will be a general tool, designed for general purposes in a general context.

If an intelligent machine were able to discern some intricate if murky regularity in data about what we have done in the past, it might be able to extrapolate about our subsequent desires, even if we don’t entirely know them ourselves.

Once a machine can translate fluently between two natural languages, the foundation has been laid for a machine that might one day understand human language well enough to engage in plausible conversation.

The story of ideas is about the cognitive scientists, psychologists and wayward engineers who long toiled in obscurity, and the process by which their ostensibly irrational convictions ultimately inspired a paradigm shift in our understanding not only of technology but also, in theory, of consciousness itself.

A beloved artifact of company culture is Jeff Dean Facts, written in the style of the Chuck Norris Facts meme: Jeff Dean’s PIN is the last four digits of pi. When Alexander Graham Bell invented the telephone, he saw a missed call from Jeff Dean. Jeff Dean got promoted to Level 11 in a system where the maximum level is 10. (This last one is, in fact, true.)

Dean was intrigued enough to lend his 20 percent — the portion of work hours every Google employee is expected to contribute to programs outside his or her core job — to the project

in the summer of 1956, a majority of researchers have long thought the best approach to creating A.I. would be to write a very big, comprehensive program that laid out both the rules of logical reasoning and sufficient knowledge of the world.

This perspective is usually called symbolic A.I. — because its definition of cognition is based on symbolic logic — or, disparagingly, good old-fashioned A.I.

languages tend to have as many exceptions as they have rules.

the proponents of symbolic A.I. took it for granted that no activities signaled general intelligence better than math and chess.

In the 1980s, a robotics researcher at Carnegie Mellon pointed out that it was easy to get computers to do adult things but nearly impossible to get them to do things a 1-year-old could do, like hold a ball or identify a cat.

a dissenting view — in which the computers would learn from the ground up (from data) rather than from the top down (from rules).

With life experience, depending on a particular person’s trials and errors, the synaptic connections among pairs of neurons get stronger or weaker.

An artificial neural network could do something similar, by gradually altering, on a guided trial-and-error basis, the numerical relationships among artificial neurons. It wouldn’t need to be preprogrammed with fixed rules. It would, instead, rewire itself to reflect patterns in the data it absorbed.

This attitude toward artificial intelligence was evolutionary rather than creationist.

You wanted to begin with very basic abilities — sensory perception and motor control — in the hope that advanced skills would emerge organically.

Google Brain’s investment allowed for the creation of artificial neural networks comparable to the brains of mice.

All they’re doing is shuffling information around in search of commonalities — basic patterns, at first, and then more complex ones — and for the moment, at least, the greatest danger is that the information we’re feeding them is biased in the first place.

our facility with what we consider the higher registers of thought are no different in kind from what we’re tempted to perceive as the lower registers. Logical reasoning, on this account, is seen as a lucky adaptation; so is the ability to throw and catch a ball. Artificial intelligence is not about building a mind; it’s about the improvement of tools to solve problems. As Corrado said to me on my very first day at Google, It’s not about what a machine ‘knows’ or ‘understands’ but what it ‘does,’ and — more importantly — what it doesn’t do yet.

What radiologists do turns out to be something much closer to predictive pattern-matching than logical analysis. They’re not telling you what caused the cancer; they’re just telling you it’s there.
december 2016 by aries1988
Soon We Won’t Program Computers. We’ll Train Them Like Dogs
For much of computing history, we have taken an inside-out view of how machines work. First we write the code, then the machine expresses it. This worldview implied plasticity, but it also suggested a kind of rules-based determinism, a sense that things are the product of their underlying instructions. Machine learning suggests the opposite, an outside-in view in which code doesn’t just determine behavior, behavior also determines code. Machines are products of the world.
programming  ai 
november 2016 by aries1988
Johnson: OK computer
How computers can be better language teachers than human beings
language  ai  tool  pc  learn 
november 2016 by aries1988
ai  tool  online  image  photo 
september 2016 by aries1988
Homo Deus by Yuval Noah Harari review – how data will destroy human freedom
human nature will be transformed in the 21st century because intelligence is uncoupling from consciousness.

Homo Deus is an ¡°end of history¡± book, but not in the crude sense that he believes things have come to a stop. Rather the opposite: things are moving so fast that it¡¯s impossible to imagine what the future might hold.
In 1800 it was possible to think meaningfully about what the world of 1900 would be like and how we might fit in. That¡¯s history: a sequence of events in which human beings play the leading part. But the world of 2100 is at present almost unimaginable.

Harari thinks the modern belief that individuals are in charge of their fate was never much more than a leap of faith. Real power always resided with networks.

The future could be a digitally supercharged version of the distant past: ancient Egypt multiplied by the power of Facebook.

Harari cares about the fate of animals in a human world but he writes about the prospects for homo sapiens in a data-driven world with a lofty insouciance.

Homo Deus makes it feel as if we are standing at the edge of a cliff after a long and arduous journey. The journey doesn’t seem so important any more. We are about to step into thin air.
human  ai  book  debate  future 
september 2016 by aries1988
Planet of the apps — have we paved the way for our own extinction? —
Harari’s skill lies in the way he tilts the prism in all these fields and looks at the world in different ways, providing fresh angles on what we thought we knew. No matter how scary and incomplete, the result is scintillating.

He points to the success of the Montreal Protocol of 1987 as a great model of international co-operation and solidarity. This treaty, ratified by 197 countries, played a vital role in reducing the release of harmful ozone-depleting chlorofluorocarbons from aerosols and refrigeration systems.

For the moment, the rise of populism, the rickety architecture of the European Union, the turmoil in the Middle East and the competing claims on the South China Sea will consume most politicians’ attention.
human  future  biology  technology  challenge  environment  book  ai  debate  crisis 
september 2016 by aries1988
Paul Taylor · The Concept of ‘Cat Face’: Machine Learning · LRB 11 August 2016
That accolade probably now belongs to move 78 in the fourth game between Sedol and AlphaGo, a moment of apparently inexplicable intuition which gave Sedol his only victory in the series. The move, quickly named the Touch of God, has captured the attention not just of fans of Go but of anyone with an interest in what differentiates human from artificial intelligence.

In 2011 he founded DeepMind with, he has said, a two-step plan to ‘solve intelligence, and then use that to solve everything else’.

One of the differences Dreyfus identified between human intelligence and digital computation is that humans interpret information in contexts that aren’t explicitly and exhaustively represented.
ai  explained  history  technology 
august 2016 by aries1988
The AI Revolution: The Road to Superintelligence
Currently, the world’s fastest supercomputer, China’s Tianhe-2, has actually beaten that number, clocking in at about 34 quadrillion cps. But Tianhe-2 is also a dick, taking up 720 square meters of space, using 24 megawatts of power (the brain runs on just 20 watts), and costing $390 million to build. Not especially applicable to wide usage, or even most commercial or industrial usage yet.

The brain’s neurons max out at around 200 Hz, while today’s microprocessors (which are much slower than they will be when we reach AGI) run at 2 GHz, or 10 million times faster than our neurons. And the brain’s internal communications, which can move at about 120 m/s, are horribly outmatched by a computer’s ability to communicate optically at the speed of light.

when it hits the lowest capacity of humanity—Nick Bostrom uses the term “the village idiot”—we’ll be like, “Oh wow, it’s like a dumb human. Cute!”
numbers  ai  comparison  brain  computer 
august 2016 by aries1988
How victory for Google’s Go AI is stoking fear in South Korea
A nation of Go players, South Korea feels Google’s victory more clearly than anyone else. How can a computer shake up a whole country?
korea  ai 
march 2016 by aries1988
The First Person to Hack the iPhone Built a Self-Driving Car. In His Garage.
The car does, more or less, have it. It stays true around the first bend. Near the end of the second, the Acura suddenly veers near an SUV to the right; I think of my soon-to-be-fatherless children; the car corrects itself. Amazed, I ask Hotz what it felt like the first time he got the car to work.

Dude, he says, the first time it worked was this morning.

There are two breakthroughs that make Hotz’s system possible. The first comes from the rise in computing power since the days of the Grand Challenge. He uses graphics chips that normally power video game consoles to process images pulled in by the car’s camera and speedy Intel chips to run his AI calculations. Where the Grand Challenge teams spent millions on their hardware and sensors, Hotz, using his winnings from hacking contests, spent a total of $50,000—the bulk of which ($30,000) was for the car itself.

Sometimes the Acura seemed to lock on to the car in front of it, or take cues around a curve from a neighboring car. Hotz hadn’t programmed any of these behaviors into the vehicle. He can’t really explain all the reasons it does what it does. It’s started making decisions on its own.

As Hotz puts it in developer parlance,  ‘If’ statements kill. They’re unreliable and imprecise in a real world full of vagaries and nuance. It’s better to teach the computer to be like a human, who constantly processes all kinds of visual clues and uses experience, to deal with the unexpected rather than teach it a hard-and-fast policy.

The truth is that work as we know it in its modern form has not been around that long, and I kind of want to use AI to abolish it. I want to take everyone’s jobs. Most people would be happy with that, especially the ones who don’t like their jobs. Let’s free them of mental tedium and push that to machines. In the next 10 years, you’ll see a big segment of the human labor force fall away. In 25 years, AI will be able to do almost everything a human can do. The last people with jobs will be AI programmers.

He thinks machines will take care of much of the work tied to producing food and other necessities. Humans will then be free to plug into their computers and get lost in virtual reality.

All this talk represents an evolution in Hotz’s hacker ethos. He used to rip apart products made by Apple and Sony, because he enjoyed solving hard puzzles and because he reveled in the thought of one person mucking up multibillion-dollar empires. With the car, the retail software, and the plans to roil entire economies, Hotz wants to build a reputation as a maker of the most profound products in the world—things that forever change how people live. I don’t care about money, he says. I want power. Not power over people, but power over nature and the destiny of technology. I just want to know how it all works.
ai  transport  car  driving  future  self  hacking 
december 2015 by aries1988
Google Just Open Sourced TensorFlow, Its Artificial Intelligence Engine | WIRED
But after they’ve been trained—when it’s time to put them into action—these neural nets run in different ways. They often run on traditional computer processors inside the data center, and in some cases, they can run on mobile phones. The Google Translate app is one mobile example. It can run entirely on a phone—without connecting to a data center across the ‘net—letting you translate foreign text into your native language even when you don’t have a good wireless signal. You can, say, point the app at a German street sign, and it will instantly translate into English.
google  opensource  ai 
november 2015 by aries1988
Software can now beat any human player at poker - New Scientist
Poker is a popular test bed for artificial intelligence research because, unlike games like chess or checkers, each player holds cards that other players cannot see. "The whole interesting part of the game comes from the fact that you don't have perfect information," says Michael Bowling, a member of the team that devised the new software.
ai  poker  game 
august 2015 by aries1988
For Sympathetic Ear, More Chinese Turn to Smartphone Program

In computer memory, spoken words, typed sentences and visual objects are represented as three-dimensional arrays of numbers that become more accurate as millions of additional images or utterances are added to the database, improving their ability to accurately recognize patterns.
august 2015 by aries1988
Chris Urmson: How a driverless car sees the road
Statistically, the least reliable part of the car is ... the driver. Chris Urmson heads up Google's driverless car program, one of several efforts to remove…
driving  explained  ai 
july 2015 by aries1988
人工智能:何时是“他们”,何时是“我们”? | 科学人 | 果壳网 科技有意思


无论是出于何种原因,借用道金斯在《自私的基因》一书中的话来说就是: “只有人类,才能够反抗基因的暴政”。
ai  human  biology  mind  enemy  distinction 
june 2015 by aries1988
Rise of the machines | The Economist
Not all future technology meets with his approval, though. In a speech in October at the Massachusetts Institute of Technology, Mr Musk described artificial intelligence (AI) as “summoning the demon”, and the creation of a rival to human intelligence as probably the biggest threat facing the world.

Much of the current excitement concerns a subfield of it called “deep learning”, a modern refinement of “machine learning”, in which computers teach themselves tasks by crunching large sets of data.

have, in the past, struggled with things that people find trivial, such as recognising faces, decoding speech and identifying objects in images.

At the same time, the most powerful computers have, in the past, struggled with things that people find trivial, such as recognising faces, decoding speech and identifying objects in images.

To take one famous example, adults can distinguish pornography from non-pornography. But describing how they do so is almost impossible, as Potter Stewart, an American Supreme Court judge, discovered in 1964.

Earlier neural networks, moreover, had only a limited appetite for data. Beyond a certain point, feeding them more information did not boost their performance. Modern systems need far less hand-holding and tweaking. They can also make good use of as many data as you are able throw at them. And because of the internet, there are plenty of data to throw.

The problem of information overload turns out to contain its own solution, especially since many of the data come helpfully pre-labelled by the people who created them. Fortified with the right algorithms, computers can use such annotated data to teach themselves to spot useful patterns, rules and categories within.

At a recent competition held at CERN, the world’s biggest particle-physics laboratory, deep-learning algorithms did a better job of spotting the signatures of subatomic particles than the software written by physicists—even though the programmers who created these algorithms had no particular knowledge of physics.

There is no result from decades of neuroscientific research to suggest that the brain is anything other than a machine, made of ordinary atoms, employing ordinary forces and obeying the ordinary laws of nature. There is no mysterious “vital spark”, in other words, that is necessary to make it go. This suggests that building an artificial brain—or even a machine that looks different from a brain but does the same sort of thing—is possible in principle.

AI uses a lot of brute force to get intelligent-seeming responses from systems that, though bigger and more powerful now than before, are no more like minds than they ever were. It does not seek to build systems that resemble biological minds.

As Edsger Dijkstra, another pioneer of AI, once remarked, asking whether a computer can think is a bit like asking “whether submarines can swim”.
ai  reportage  explained 
may 2015 by aries1988
Alex Garland of ‘Ex Machina’ Talks About Artificial Intelligence

But there is a mistake here. The machines in question are not strong A.I.’s. They are weak. They have no motivation, no intention; they’re neutral. The thing with an agenda is us: consumers, who want to buy the machines, and manufacturers, who want to sell them. And looming over both, giant tech companies, whose growth only ever seems to be exponential, whose practices are opaque, and whose power is both massive and without true oversight. Combine all this with government surveillance and lotus-eating public acquiescence, and it’s not the machine component that scares me. It’s the human component.
movie  ai  future  human  society  industry  essay 
april 2015 by aries1988
The Machines Are Coming

To crack these cognitive and emotional puzzles, computers needed not only sophisticated, efficient algorithms, but also vast amounts of human-generated data, which can now be easily harvested from our digitized world. The results are dazzling. Most of what we think of as expertise, knowledge and intuition is being deconstructed and recreated as an algorithmic competency, fueled by big data.

Technology in the workplace is as much about power and control as it is about productivity and efficiency.

We don’t need to reject or blame technology. This problem is not us versus the machines, but between us, as humans, and how we value one another.
workforce  ai  crisis  opinion 
april 2015 by aries1988
The End Is A.I.: The Singularity Is Sci-Fi's Faith-Based Initiative
Is machine sentience not only possible, but inevitable? Of course not. But don't tell that to devotees of the Singularity, a sci-fi-inspired theory that sounds like science, but is really just the rapture for nerds.

This is what Vinge dubbed the Singularity, a point in our collective future that will be utterly, and unknowably transformed by technology’s rapid pace. The Singularity—which Vinge explores in depth, but humbly sources back to the pioneering mathematician John von Neumann—is the futurist’s equivalent of a black hole, describing the way in which progress itself will continue to speed up, moving more quickly the closer it gets to the dawn of machine super intelligence. Once artificial intelligence (AI) is accomplished, the global transformation could take years, or mere hours. Notably, Vinge cites a SF short story by Greg Bear as an example of the latter outcome, like a prophet bolstering his argument for the coming end-times with passages from scripture.

Neuroscience has made incredible progress in directly observing the brain, and drawing connections between electrical and neurochemical activity and some forms of behavior. But the closer we zoom in on the mind, the more complex its structures and patterns appear to be. The purpose of the Human Brain Project, as well as the United States’s BRAIN Initiative, is to address the fact that we know astonishingly little about how and why human beings think about anything. These projects aren’t signs of triumph over our biology, or indications that the finish line is near. They’re admissions of humility.

“If you asked someone, 50 years ago, what the first computer to beat a human at chess would look like, they would imagine a general AI,” says Naam. “It would be a sentient AI that could also write poetry and have a conception of right and wrong. And it’s not. It’s nothing like that at all.” Though it outplayed Garry Kasparov in 1997, IBM’s Deep Blue is no closer to sentience than a ThinkPad laptop. And despite its Jeopardy prowess, “IBM’s Watson can’t play chess, or drive one of Google’s robotic cars,” says Naam. “We’re not actually trending towards general AI in anyway. We’re just building better and better specialized systems.”
scifi  ai  opinion  future  brain 
december 2014 by aries1988
« earlier      
per page:    204080120160

Copy this bookmark:

to read