Artificial General Intelligence and the tech singularity threat: 11 AI insights from experts

What used to be labeled as science-fiction few decades ago may now be a forthcoming reality, except for the whole ‘Terminator’ craze.

Artificial Intelligence research goes back to the late fifties, where the link between human intelligence and machines was widely observed. Fast forward to today, human intelligence is facing a real threat.

In fact, machines are getting twice better every 18 months and we’re now increasingly hearing of Artificial General Intelligence (AGI), when machines start exhibiting intelligence that exceed human performance in all domains, versus Artificial Narrow Intelligence (ANI), which is when machines outperform humans in certain domains. We’re currently in the ANI phase.

Are humans on the verge of becoming the second most intelligent species on earth?

In order to better understand AGI, how can we prevent it from destroying us and knowing its implications on several fields and economies, The Order of Engineers and Architects (OAE) in Beirut organized a two-day event on February 26 and 27, 2019 called Artificial Intelligence, Digital Revolution & Impact on the Economy.

The event included several talks and two panels and featured speakers from engineering background, as well as data analysis, AI, IoT, fintech, politics, economy, philosophy and others. We were also excited to have four of our LebNet members giving talks or moderating sessions.

The content of the sessions was highly rich and informative, backed with statistics, numbers and real examples. Below are some of the insights and observations that experts shared with us during the event. While they are not representative of the event, they do give a general idea of what has been discussed:

1- If tech progress was not combined with the preservation of good values and acting civilized, we will be facing a social deterioration, with one minority taking over politics and economy and the majority being marginalized. Quoting Einstein: It has become appallingly clear that our tech has exceeded our humanity. Jad Tabet, the President of the OEA.

2- People are debating as to when we will reach AGI. Some experts say as early as 2029, other say it might take 70 years. When we hit AGI, singularity will happen and humans will no longer be the most intelligent species on this planet. Mazen Skaf, Partner and Managing Director of Strategic Decisions Group.

3- AI crushed us in 4 competitions: the first one in 1997, when the world’s best chess player, Gary Kasparov, was beaten by a computer; the second in 2011, when IBM’s supercomputer Watson beat Ken Jennings during a TV quiz show; the third in 2016, when Google DeepMind’s AlphaGo algorithm beat the world’s number one Go player and the fourth in 2017, when the best poker players were beaten by a computer. Calum Chase, Author of ‘Surviving AI’ and ‘The Economic Singularity’.

4- AI is a misunderstood word. Everybody uses and misuses it. Machine learning is reversing the way we write programs and neural networks is inspired by the human brain. It has multiple layers and each layer is creating the features that will be used by machine learning. Bassem Monla, AI Subject Matter Expert at IBM.

5- Estimates by Mckinsey suggest that by 2030, we will be adding about 13 trillion dollars to the economy, thanks to automation and optimizing productivity. The countries that are most likely to benefit from that are the countries that are already developed. Today US and China have the highest level of AI skills penetration. Nasser Saidi, Economist and Former Minister of Economy in Lebanon.

6- I advise Lebanon to do a leapfrog. We can skip a generation of technology and get up to speed and become competitive by skipping a lot of technology that has been around for the last 10 to 20 years. Imad ElHajj, Professor at the American University of Beirut.

7- Computers can learn and understand feelings. We collected 120 pieces of music and published them online and asked people to listen to the music and tell us what emotions the music evoked in them. We collected the feedback and used the data to train our sentiment learner to detect and analyse emotions from the music and compose music on its own. Joe Tekli, Assistant Professor/Interim Assistant Dean/School of Engineering at the Lebanese American University.

8- In 2019, we have 130 devices being connected to the internet every second. Analysts predictions say we’re getting to 75 billion devices connecting to the internet by 2025. Rabih Nassar, CEO of Scriptr

9- In China, banks use AI to look at your social media accounts, your web browsing history and chats, then evaluate your credit worthiness. Convolutional Neural Network (CNN) algorithm is very much used today to recognise images and signature fraud used in banks. Next in Natural Language Processing we have chatbots. In wealth management it’s very well done. It prepares the room for analyst to be able to give you a better service. Crédit Mutuel, is using Watson and has improved customer satisfaction by 60% in wealth management. Chatbots, When they understand you it’s perfect, when they don’t it’s a catastrophe. Gerard Rafie, founder of FintekMinds.

10- The first AGI machines that we create should like us and understand us better than we understand ourselves. In the next couple of decades, people will have to re-skill and retrain themselves to new jobs. If AI takes over our jobs, we could focus on other things: become the best mountaineer or painter. We can have a second renaissance. And if we make the cost of living very cheap with a Universal Basic Income for all and rely on economic abundance, we can achieve that. Callum Chase.

11- AI might help us save ourselves from the problems we created: climate change, sustainability, mass inequality and cyberattacks. Some companies are using AI to identify cyberattacks from day 0 and intervene within seconds. The future of AI lies on adopting a code of ethics in the industry. Mazen Skaf.

Taking into account that it’s not easy to host several sessions discussing one major topic without being repetitive, The OEA’s event did a good job providing rich and varied content in every talk. We would like to thank our four LebNet members, Bassem Monla, Rania Afiouni, Gerard Rafie and Mazen Skaff for making the trip from Montreal, Austin and California to join the event.