Super intelligents AI-Life 3.0

Interview: Max Tegmark on Superintelligent AI, Cosmic Apocalypse, and Life 3.0

Image: Penguin Random House

Ask Max Tegmark why people should read his new book and get involved in the discussion about artificial intelligence, and you get a weighty answer. Forget about book sales, this is about cosmic destiny: The fate of the universe may well be determined by the decisions made “here on our little planet during our lifetime,” he says.

In his book, Life 3.0: Being Human in the Age of Artificial Intelligence, Tegmark first explains how today’s AI research will likely lead to the creation of a superintelligent AI, then goes further to explore the possible futures that could result from this creation. It’s not all doom and gloom. But in his worst case scenario, humanity goes extinct and is replaced with AI that has plenty of intelligence, but no consciousness. If all the wonders of the cosmos carry on without a conscious mind to appreciate them, the universe will be rendered a meaningless “waste of space,” Tegmark argues.

Tegmark, an MIT physics professor, has emerged as a leading advocate for research on AI safety. His thoughtful book builds on the work of Nick Bostrom, who famously freaked out Elon Musk with his book Superintelligence, which described in meticulous detail how a supercharged AI could lead to humanity’s destruction.

Max Tegmark on . . .

  1. Why He Disagrees With Yann LeCun
  2. “I Really Don’t Like It When People Ask What Will Happen in the Future”
  3. What Types of AI Safety Research We Should Fund Now
  4. The Question of Consciousness
  5. Cosmic Optimism vs Cosmic Pessimism
  6. AI as the “Child of All Humanity”

Superintelligent AI, Cosmic Apocalypse, and Life 3.0

Will Expert System Come To Be Conscious?

Ignore today’s small step-by-step advances in artificial intelligence, such as the boosting capacities of cars and trucks to drive themselves. Waiting in the wings could be a groundbreaking growth: a machine that knows itself as well as its environments, and that might take in and also procedure massive amounts of data in genuine time. Maybe sent out on harmful objectives, into space or combat. In addition to driving individuals about, it could be able to prepare, clean, do washing– as well as keep people business when other people typically aren’t close by.

A specifically innovative set of machines might replace humans at essentially all tasks. That would certainly conserve humankind from workaday drudgery, however it would certainly also tremble several social foundations. A life of no job and also just play may end up being a dystopia.

Aware makers would certainly also increase uncomfortable lawful and moral problems. Would certainly a mindful machine be a “individual” under regulation and also be responsible if its actions injure a person, or if something fails? To think of a more frightening circumstance, might these devices rebel versus people and desire to remove us entirely? If yes, they represent the culmination of evolution.

As a teacher of electric engineering and also computer technology who operates in machine learning and quantum theory, I could say that scientists are divided on whether these sorts of hyperaware makers will ever exist. There’s additionally question regarding whether devices might or must be called “aware” in the way we think of human beings, and even some animals, as conscious. Several of the questions relate to modern technology; others concern exactly what consciousness in fact is.

Is Understanding Sufficient?
The majority of computer system researchers think that awareness is a characteristic that will emerge as modern technology develops. Some think that consciousness includes approving new info, saving and fetching old information as well as cognitive handling of everything right into assumptions and activities. If that’s right, then one day makers will without a doubt be the supreme awareness. They’ll have the ability to gather even more information than a human, store more than many collections, accessibility huge databases in nanoseconds as well as compute all of it into choices extra complicated, and yet much more logical, than anyone ever could.

On the other hand, there are physicists and also theorists who say there’s something extra about human habits that can not be computed by an equipment. Creative thinking, as an example, and the feeling of liberty individuals possess don’t appear to find from reasoning or estimations.

Yet these are not the only views of what awareness is, or whether makers could ever achieve it.

Quantum Sights
One more viewpoint on awareness originates from quantum theory, which is the inmost concept of physics. Inning accordance with the orthodox Copenhagen Interpretation, consciousness and also the physical world are complementary facets of the same fact. When a person observes, or experiments on, some aspect of the physical world, that individual’s mindful communication triggers noticeable change. Because it takes consciousness as a provided as well as no attempt is made to acquire it from physics, the Copenhagen Analysis might be called the “big-C” sight of awareness, where it is a point that exists on its own– although it needs minds to become actual. This sight was popular with the leaders of quantum theory such as Niels Bohr, Werner Heisenberg and Erwin Schrodinger.

The communication in between consciousness and also matter brings about mysteries that continue to be unsettled after 80 years of debate. A widely known instance of this is the paradox of Schrodinger’s pet cat, in which a feline is positioned in a scenario that leads to it being similarly most likely to survive or pass away– and also the act of observation itself is exactly what makes the outcome specific.

The opposing sight is that consciousness arises from biology, just as biology itself emerges from chemistry which, in turn, emerges from physics. We call this less large concept of consciousness “little-C.” It agrees with the neuroscientists’ view that the procedures of the mind are identical to states as well as procedures of the mind. It also agrees with a much more recent analysis of quantum theory inspired by an effort to rid it of mysteries, the Many Worlds Interpretation, in which viewers belong of the mathematics of physics.

Philosophers of scientific research believe that these modern quantum physics views of consciousness have parallels in old approach. Big-C resembles the concept of mind in Vedanta– in which consciousness is the essential basis of reality, on the same level with the physical world.

Little-C, in contrast, is rather much like Buddhism. Although the Buddha chose not to resolve the concern of the nature of awareness, his fans declared that mind and consciousness develop from emptiness or nothingness.

Big-C and also Scientific Exploration
Researchers are also exploring whether consciousness is always a computational procedure. Some scholars have argued that the innovative minute is not at the end of a purposeful calculation. For example, dreams or visions are supposed to have motivated Elias Howe’s 1845 design of the modern stitching maker, and August Kekule’s exploration of the framework of benzene in 1862.

A dramatic item of evidence in favor of big-C awareness existing all on its own is the life of self-taught Indian mathematician Srinivasa Ramanujan, who passed away in 1920 at the age of 32. His notebook, which was lost and neglected for about 50 years and released just in 1988, consists of a number of thousand formulas, without evidence in various areas of math, that were well in advance of their time. Moreover, the methods whereby he located the solutions remain elusive. He himself declared that they were exposed to him by a goddess while he was asleep.

The idea of big-C awareness increases the questions of exactly how it relates to matter, and exactly how matter as well as mind equally affect each other. Consciousness alone can not make physical modifications to the world, but probably it could transform the possibilities in the evolution of quantum processes. The act of observation could freeze as well as influence atoms’ movements, as Cornell physicists confirmed in 2015. This might quite possibly be a description of just how matter as well as mind connect.

Mind and also Self-Organizing Equipments
It is possible that the phenomenon of awareness needs a self-organizing system, like the mind’s physical framework. If so, then existing machines will certainly lose.

Scholars don’t know if adaptive self-organizing equipments can be created to be as advanced as the human mind; we lack a mathematical concept of calculation for systems like that. Perhaps it’s true that just organic machines can be sufficiently imaginative and adaptable. However then that suggests people need to– or soon will– begin working with engineering brand-new organic structures that are, or could end up being, aware.

Emotionally Intelligent AI

The Rise of Emotionally Intelligent AI

I have studied emotional intelligence as a hobby for a long time. Until recently, I believed emotional intelligence to remain one of the core advantages of us humans after artificial intelligence has taken over all tasks requiring memorization and logic.

During the past few years, I’ve focused my studies on emotionally intelligent algorithms, as it is the business of my startup, Inbot.

The more I have researched them, the more convinced I have become that people are no longer ahead of AI at emotional intelligence.

Yuval Noah Harari writes in his best-selling book Homo Deus that humans are essentially a collection of biological algorithms shaped by millions of years of evolution. He continues to claim that there is no reason to think that non-organic algorithms couldn’t replicate and surpass everything that organic algorithms can do.

The same is echoed by Max Tegmark in his book Life 3.0: Being Human in the Age of Artificial Intelligence. He makes a compelling case that practically all intelligence is substrate independent.

Let that sink in for a moment. Our emotions and feelings are organic algorithms that respond to our environment. Algorithms, that are shaped by our cultural history, upbringing and life experiences. And they can be reverse engineered.

If we agree with Dr. Harari, who is a professor at the Hebrew University of Jerusalem, and Dr. Tegmark, who is a professor at MIT in Boston, computers will eventually become better at manipulating human emotions than humans themselves.

People are generally not emotionally intelligent

In real life situations, we are actually pretty bad at emotional intelligence.

Most of us are ignorant about even the most basic emotional triggers we set off in others. We end up in pointless fights, dismiss good arguments because they go against our biases, and judge people based on stereotypes.

We don’t understand the effects of cultural context, family upbringing or the current personal life situation of our discussion partner.

We rarely try to put ourselves in the other person’s position. We don’t try to understand their reasoning if it goes against our worldview. We don’t want to challenge our biases or prejudices.

Online, the situation is much worse. We draw hasty and often mistaken conclusions from comments by people we don’t know at all, and lash at them if we believe their point goes against our biases.

Lastly, we have an evolutionary trait of seeing life as the “survival of the fittest”. This predisposes us from taking advantage of others, to focus on boosting our egos, and to put ourselves on a pedestal.

The most successful people often lie to gain advantage, manipulate to get ahead, and deceive to hide their wrongdoings. It’s about winning at all costs, causing a lot of emotional damage on the way.

AI is advancing rapidly at emotional intelligence

While us humans continue to struggle to understand each other, emotionally intelligent AI has advanced rapidly.

Cameras in phones are ubiquitous and omnipresent, and face-tracking software is already advanced enough to analyze the smallest details of our facial expressions. The most advanced ones can even tell apart faked emotions from real ones.

In addition, voice recognition and natural language processing algorithms are getting better at figuring out our sentiment and emotional state from the audio.

The technologies to analyze emotional responses from faces and voice are already way beyond the skills of an average human, and in many areas exceed the abilities of even the most skilled humans.

Artificial Intelligence can look at our faces to recognize such private qualities as your sexual orientation, political leaning or IQ.

While AI can decipher almost any emotion from your face or speech, we haven’t yet put a lot of effort in scientific study of emotionally intelligent AIs.

The advances in this field are currently almost solely driven by commercial interests and human greed.

Media and entertainment companies need our attention and engagement to make money. Companies like Facebook and YouTube have a large number of engineers working to create ever better ways to addict us to their content.

I wrote about this in earlier in a short post named The worrying growth of the business of addiction.

These algorithms are designed to pull our emotional triggers to keep us entertained. And they have become very, very good at it.

Some of the core developers of these algorithms have gotten scared of the power technology has on us, and say our minds can be hijacked.

Big data gives an edge to emotionally intelligent AIs

Unlike people, AI can leverage your whole online history, which in most cases is more information than anybody can remember about any of their friends.

Some of the most advanced machine learning algorithms developed at Facebook and Google have already been applied on a treasure trove of data from billions of people.

These algorithms already know what your desires, biases and emotional triggers are, based on your communication, friends and cultural context. In many areas, they understand you better than you know yourself.

The progress of algorithms has gone so far that Facebook and Google are now accused of creating filter bubbles that can effect public opinion, rapidly change political landscapes and sway elections.

These algorithms are getting so complex that they are becoming impossible to fully control by humans. Facebook’s chief of security Alex Stamos recently tweeted that journalists are unfairly accusing them for manipulation, when in reality there are no solutions available that wouldn’t lead to someone accusing them of bias.

The future of emotional artificial intelligence

People have a lot of biases, which cloud our judgment. We see the world as we wish it to be, not as it is. Algorithms today, being made by people, incorporate some hints of our biases too. But if we wanted to remove such biases, it would be relatively easy to do.

As artificial intelligence gets better at manipulating us, I see a future where people happily submit their lives to the algorithms. We can already see it in practice. Just look around yourself in public — almost everyone is glued to the their smartphones.

Today, people touch their phones on average 2,617 times a day.

We are approaching an era, when artificial intelligence uses humans as organic robots to realize its goals. To make that happen, thousands of engineers are already building an API to humans.

The second part of this series is called The Human API.

Berlin, 9.10.2017

Mikko Alasaarela

I look forward to debating this interesting topic with you. Please comment and share!

My company Inbot is among the pioneers that leverage AI algorithms to offer real long term monetary value to humans for their data and services. We exist to counter the trend of intelligent machines enslaving humans, and to provide human opportunity in the age of artificial intelligence.

Join our Ambassador Community to earn long term dividends by introducing innovative AI businesses to customers.


Join 30,000+ people who read the weekly ?Machine Learnings? newsletter to understand how AI will impact the way they work.

5-year trends in artificially intelligent marketing

How will artificial intelligence transform marketing over the coming years? Columnist Daniel Faggella dives into the results from a survey exploring the major trends and opportunities in AI for marketers.

Artificial intelligence has been making headlines over the last 12 months in domains like health care, finance, face recognition and more. Marketing, however, doesn’t seem to be getting the same kind of coverage, despite major developments in the application of AI to marketing analytics and business intelligence.

Five or 10 years ago, only the world’s savviest, most heavily funded companies had a serious foothold in artificial intelligence marketing tech. As we enter 2017, there are hundreds of AI marketing companies all over the world (including some that have gone public, like RocketFuel). These companies are making AI and machine learning accessible to large corporations and SMBs (small and medium-sized businesses) alike, opening new opportunities for smarter marketing decisions and approaches.

Over the last three months, we surveyed over 50 machine learning marketing executives (email registration required for the full report data) to get a sense of the important trends and implications of AI over the next five years.

Below, I’ve highlighted three major trends that impact the theme of “Intelligent Content.”

Recommendation and personalization predicted to be greatest profit opportunity

While most of our executives voted “Search” as the AI marketing tool with the highest profit potential today, “Recommendation and Personalization” topped the list for ROI potential in the coming five years.

While search requires users to express their intent in text (or speech), recommendation pulls from myriad points of data and behavior — often bringing a user to a) what they were truly looking for, or b) what the advertiser wanted them to find.

The implications of recommendation in content marketing are numerous. Below I’ll list just a few:

First, recommendation engines help serve the content most likely to engage readers. In the past, this was done with simple text analysis or tools like elastic search. The “recommended” content was better than a random guess, but it was by no means truly optimized for user engagement.

Companies like Boomtrain and Liftigniter are developing technologies to tailor content to individual visitors, displaying material most likely to keep them on the site based on their previous engagements, purchases, clicks and more.

Second, programmatic advertising (like that used on giant platforms like Facebook and Google AdWords) is often used to drive users directly to content before seeing a product page or being asked to book an appointment. Many ad networks (Facebook included) don’t allow for direct lead generation and instead prefer to engage users with the right content first before looking for a conversion.

Ad networks are partial to keeping user experience high in addition to driving engagement on ads, which is a delicate balance. Companies that can leverage these programmatic platforms to target the right prospects with the right content are the most likely to win.

Third, we see entire content marketing platforms at the heart of business models. One such example is Houzz.com, a site that hosts millions of articles and photo albums about home improvement and decoration. This content ecosystem links to and references millions of home goods products (from throw rugs to couches and more), and “recommendation” drives the entire experience.

Houzz is one of the best current examples of “intelligence content” directly tying to sales, and I suspect that in the coming five years, we’ll see elements of their business model become much more prevalent.

Intelligent content might be content that makes itself

Content generation is a complex machine learning problem, and until recently, it’s been relegated to big-budget media firms working in quantitatively oriented domains (namely sports and finance). Yahoo Finance uses natural language generation (NLG) to turn information about stocks and bonds into coherent, human-readable articles, saving time for Yahoo’s writers so that they can complete more important and creative tasks.

NLG is now being used in a vast number of business applications including compliance, insurance and more — and a quick visit to the “solutions” page at Narrative Science shows a plethora of use cases and case studies for machine-written content.

While domains like finance and compliance involve strict, formulaic transformations of cold data into readable text, executives in the field are excited about its profit potential, too. Rather than simply saving costs on human writers, intelligent content generation will alter existing content (and/or create new content) to help driving marketing goals. As Laura Pressman, manager of Automated Insights, explained in our survey:

Content generation has high profit potential in the coming five years. Personalization and segmentation can be achieved through altering the content text to speak to certain groups of people, across different platforms, highlighting unique and targeted features that are most important to each specific segment.

B2C companies may have an advantage in intelligent content

When we polled our batch of executives about the most meaningful applications of artificial intelligence in marketing, we didn’t want to leave out their opinions about which businesses or industries would be most able to take advantage of AI’s advancements in marketing.

“Industry” didn’t seem to have much to do with the predicted success that a company might have with AI marketing tech. Much more important was the way the company did business and sold products. For a business to take advantage of AI, the most important traits (as predicted by our batch of executives) include:

  • Data collection: Ability to quantify customer touch points across all marketing activities.
  • Transaction volume: Reaching the marketing “goal” more often helps to train marketing algorithms and provide better predictions and recommendations.
  • Uniformity: Businesses that pool their marketing and sales data into a single stream are more likely to succeed in applying AI.

The above three qualities repeated themselves again and again in our survey responses, along with strong predictions that “Digital Media” companies and “E-commerce/Consumer Retail” companies would be most poised to take advantage of AI in marketing. As Lisa Burton, chief data officer of AdMass, explained in the survey:

Advertisers and e-commerce businesses have the highest potential gain from machine learning because of the ease of measurement and quick feedback needed to train and improve machine learning algorithms.

While B2C and retail companies seem to have an edge on “quantifiability” and attribution to sale, some of our respondents also hinted at the strong opportunity in B2B. Leveraging the many content and interaction touch points in a B2B sale will aid greatly in “cracking the code” on B2B marketing attribution, which is undoubtedly valuable.

In the coming five years, it may be possible that attribution and recommendation take off quickly in retail, while adoption in services and B2B sectors will provide more of an “ahead of the curve” advantage in industries where tech adoption is slower.

5-year trends in artificially intelligent marketing

 

A.I. will replace half of all jobs in the next decade

  • Intelligent Assistant

A Different Kind of Self Service

Tech Optimists See a Golden Future—Let’s Talk About How We’ll Get There

Technology evangelists dream about a future where we’re all liberated from the more mundane aspects of our jobs by artificial intelligence. Other futurists go further, imagining AI will enable us to become superhuman, enhancing our intelligence, abandoning our mortal bodies, and uploading ourselves to the cloud.

Paradise is all very well, although your mileage may vary on whether these scenarios are realistic or desirable. The real question is, how do we get there?

Economist John Maynard Keynes notably argued in favor of active intervention when an economic crisis hits, rather than waiting for the markets to settle down to a more healthy equilibrium in the long run. His rebuttal to critics was, “In the long run, we are all dead.” After all, if it takes 50 years of upheaval and economic chaos for things to return to normality, there has been an immense amount of human suffering first.

Similar problems arise with the transition to a world where AI is intimately involved in our lives. In the long term, automation of labor might benefit the human species immensely. But in the short term, it has all kinds of potential pitfalls, especially in exacerbating inequality within societies where AI takes on a larger role. A new report from the Institute for Public Policy Research has deep concerns about the future of work.

Uneven Distribution

While the report doesn’t foresee the same gloom and doom of mass unemployment that other commentators have considered, the concern is that the gains in productivity and economic benefits from AI will be unevenly distributed. In the UK, jobs that account for £290 billion worth of wages in today’s economy could potentially be automated with current technology. But these are disproportionately jobs held by people who are already suffering from social inequality.

Low-wage jobs are five times more likely to be automated than high-wage jobs. A greater proportion of jobs held by women are likely to be automated. The solution that’s often suggested is that people should simply “retrain”; but if no funding or assistance is provided, this burden is too much to bear. You can’t expect people to seamlessly transition from driving taxis to writing self-driving car software without help. As we have already seen, inequality is exacerbated when jobs that don’t require advanced education (even if they require a great deal of technical skill) are the first to go.

No Room for Beginners

Optimists say algorithms won’t replace humans, but will instead liberate us from the dull parts of our jobs. Lawyers used to have to spend hours trawling through case law to find legal precedents; now AI can identify the most relevant documents for them. Doctors no longer need to look through endless scans and perform diagnostic tests; machines can do this, leaving the decision-making to humans. This boosts productivity and provides invaluable tools for workers.

But there are issues with this rosy picture. If humans need to do less work, the economic incentive is for the boss to reduce their hours. Some of these “dull, routine” parts of the job were traditionally how people getting into the field learned the ropes: paralegals used to look through case law, but AI may render them obsolete. Even in the field of journalism, there’s now software that will rewrite press releases for publication, traditionally something close to an entry-level task. If there are no entry-level jobs, or if entry-level now requires years of training, the result is to exacerbate inequality and reduce social mobility.

Automating Our Biases

The adoption of algorithms into employment has already had negative impacts on equality. Cathy O’Neil, mathematics PhD from Harvard, raises these concerns in her excellent book Weapons of Math Destruction. She notes that algorithms designed by humans often encode the biases of that society, whether they’re racial or based on gender and sexuality.

Google’s search engine advertises more executive-level jobs to users it thinks are male. AI programs predict that black offenders are more likely to re-offend than white offenders; they receive correspondingly longer sentences. It needn’t necessarily be that bias has been actively programmed; perhaps the algorithms just learn from historical data, but this means they will perpetuate historical inequalities.

Take candidate-screening software HireVue, used by many major corporations to assess new employees. It analyzes “verbal and non-verbal cues” of candidates, comparing them to employees that historically did well. Either way, according to Cathy O’Neil, they are “using people’s fear and trust of mathematics to prevent them from asking questions.” With no transparency or understanding of how the algorithm generates its results, and no consensus over who’s responsible for the results, discrimination can occur automatically, on a massive scale.

Combine this with other demographic trends. In rich countries, people are living longer. An increasing burden will be placed on a shrinking tax base to support that elderly population. A recent study said that due to the accumulation of wealth in older generations, millennials stand to inherit more than any previous generation, but it won’t happen until they’re in their 60s. Meanwhile, those with savings and capital will benefit as the economy shifts: the stock market and GDP will grow, but wages and equality will fall, a situation that favors people who are already wealthy.

Even in the most dramatic AI scenarios, inequality is exacerbated. If someone develops a general intelligence that’s near-human or super-human, and they manage to control and monopolize it, they instantly become immensely wealthy and powerful. If the glorious technological future that Silicon Valley enthusiasts dream about is only going to serve to make the growing gaps wider and strengthen existing unfair power structures, is it something worth striving for?

What Makes a Utopia?

We urgently need to redefine our notion of progress. Philosophers worry about an AI that is misaligned—the things it seeks to maximize are not the things we want maximized. At the same time, we measure the development of our countries by GDP, not the quality of life of workers or the equality of opportunity in the society. Growing wealth with increased inequality is not progress.

Some people will take the position that there are always winners and losers in society, and that any attempt to redress the inequalities of our society will stifle economic growth and leave everyone worse off. Some will see this as an argument for a new economic model, based around universal basic income. Any moves towards this will need to take care that it’s affordable, sustainable, and doesn’t lead towards an entrenched two-tier society.

Walter Schiedel’s book The Great Leveller is a huge survey of inequality across all of human history, from the 21st century to prehistoric cave-dwellers. He argues that only revolutions, wars, and other catastrophes have historically reduced inequality: a perfect example is the Black Death in Europe, which (by reducing the population and therefore the labor supply that was available) increased wages and reduced inequality. Meanwhile, our solution to the financial crisis of 2007-8 may have only made the problem worse.

But in a world of nuclear weapons, of biowarfare, of cyberwarfare—a world of unprecedented, complex, distributed threats—the consequences of these “safety valves” could be worse than ever before. Inequality increases the risk of global catastrophe, and global catastrophes could scupper any progress towards the techno-utopia that the utopians dream of. And a society with entrenched inequality is no utopia at all.

Image Credit: OliveTree / Shutterstock.com

<

p class=”wpematico_credit”>Powered by WPeMatico