Brains vs computers?
Computers are incredibly fast, accurate, and stupid.
Human beings are incredibly slow, inaccurate, and brilliant.
Together they are powerful beyond imagination.” Quote attributed to Albert Einstein.
This article aims at introducing the ambiguous relationship between human brains and modern computers.
Brains vs computers
How different are they? How can one help another?
Although they both use electrical signals to send messages, our nervous system is far from being as fast as the computer’s wires. Our memory can grow, with practice, as our synaptic connections get stronger but the computer’s short memory (RAM) can be easily updated and it increases its performance as well. Today, PC’s long-term memory can store a huge among of data compared to our fragile encephalon where our memories quickly fade away and get altered with time. The learning process is another source of dissimilarity. Computers are known for their ability to perform multiple tasks simultaneously. All of today’s operating systems (Windows, Mac OS, Linux…) can easily let the user open a web page via his web browser and prepare his upcoming presentation on power point [1]. Multitasking is, on the contrary, extremely complex for human brains [2].
Brains & computers
Towards a humanization of computers?…
Artificial intelligence, Deep Learning, Intelligent Systems and many other applications are essentially trying to humanize computers so that the latter can take over tasks that we do on a daily basis like transforming speech into words, recognizing faces or even lying.[3]
IBM Research Vice President and Lab Director Jeffrey Welser mentioned, during a keynote presentation at Stanford, this humanization process. According to him, software neural networks[4], which are composed of numerous deep learning technologies, are “approaching human accuracy”. Since the introduction of large scale neural networks in both the Image recognition and Speech recognition processes (in 2012 and 2010 respectively), the error rate has dropped significantly. Actually, computers now have a smaller percentage of error (5-6%) than human beings on average (between 5 and 11%) when it comes to image recognition. Regarding Speech recognition, computers are approaching human accuracy.
…Or a computerized brain?
Companies and entrepreneurs like Elon Musk (yes, him again), through his new company Neuralink, try to reverse the process and actually computerize our human brains through electrode arrays implanted in the brain. According to him, “a merger of biological intelligence and machine intelligence would be necessary for humans to compete with machines”. Such technology could allow us to “upload and download thoughts”[5] and unlock our potential to learn, store data and many other things that our limited abilities prevents us from doing. Some would even suggest the possibility of uploading someone’s mind into a computer although the complexity of the human mind remains an unsolved mystery as of today. On this topic, Ray Kurzweil, director of engineering at Google recently stated that our species was going to become “increasingly non-biological to the point where the non-biological part dominates and the biological part is not important anymore.”
Therefore, as the line between human and computers gets thinner and thinner, ethical issues must be raised.
As it is the case with geoengineering or even gene-editing technologies, limitations must be created and implemented. With the advances in AI, philosophers like Daniel Dennet and John Sullins question the potential moral responsibility of computers. As they will progressively act with interests and a purpose, will that make them moral agents? Sullins believes moral agency “does not require personhood”[6]. According to him, as long as the computer shows intentionality in its behavior, it deserves moral consideration. One could also say that computers do not currently possess the necessary level of abstraction to reflect on their beliefs and intentions; and that this “extra-layer” of consciousness could separate moral agents from the rest.
References:
1. Human beings can actually perform many tasks at the same time using the autonomous nervous system. The latter regulates numerous body processes such as the blood pressure, the heart rate and the body temperatures among many others. To know more: http://www.merckmanuals.com/home/brain,-spinal-cord,-and-nerve-disorders/autonomic-nervous-system-disorders/overview-of-the-autonomic-nervous-system
2. Source on computers multitasking ability: http://whatis.techtarget.com/definition/multitasking
3. Source on lying: https://www.theregister.co.uk/2017/06/15/facebook_to_teach_chatbots_negotiation/
4. To know more about AI and neural networks, this report is a good start: http://searchnetworking.techtarget.com/definition/neural-network
5. Source: https://www.wsj.com/articles/elon-musk-launches-neuralink-to-connect-brains-with-computers-1490642652
6. Source: https://plato.stanford.edu/entries/computing-responsibility/
Users who have LIKED this post:
7 comments on “Brains vs computers?”
Comments are closed.
Hello Victor. Thank you for sharing.
With your article, two pieces of work came to my mind: “Ghost in the Shell(G.I.T.S.)” and “Childhood’s End” by Sir Arthur C. Clarke. If you haven’t seen, you might be intrigued.
FYI, as for G.I.T.S., I recommend original animation work, rather than Scarlett Johansson.
https://www.youtube.com/watch?v=p2MEaROKjaE
Users who have LIKED this comment:
Hey Victor, really interesting article and thanks for sharing! The rapid development of AI indeed raised ethical questions that people cannot ignore in this era when standing from a humanity perspective. The line of human brain and computer is definitely not only getting thinner but also more ambiguous. In the case of “Ghost in the Shell”, the plot is developed with a setting that several scientists turned memories and thoughts of a person into the form of data and then stored it inside another person’s brain. Although that story was written as a fiction, I can totally see how it can be realized with Elon Mask’s vision on Neuralink. Actually, if you are interested, there’s course currently at Stanford called PHIL29S: Philosophy and Emerging Technologies that I thought is kind of related with what you are discussing here.
However, my other concerns are from a business perspective. Although Elon Mask’s vision seems pretty real and has strong technical support, I highly doubt the speed of consumers’ adaptation on that. AI is a technology that’s in rapid development but has not yet been highly consumerized. We have seen various demos and the so-called buzzwords among the circulation of engineers, but it has not yet been fully integrated into real life because few people outside the industry understand the real meaning of it. Therefore, if the emergence of a product requires a huge shift of everyday user habit, it may encounter an early phase of unacceptance among normal consumers.
Users who have LIKED this comment:
Hey Victor,
Very interesting contrast between brains and computers. One thing that I found interesting was the evolution of AI and Deep Learning and how you were able to demonstrate its progression over time. What I’d like to learn about is how companies are looking to improve the creativity aspect of computers and how we can develop computers to help them become more than a tool. I find the viewpoints you referenced that were hypothesized by Musk and Kruzwell concerning from a social perspective as it does raise some serious ethical issues and opens up humans to a whole host of vulnerabilities that affect computers today. I guess what it really comes to are tradeoffs; at what point does the digitization and computerization of the human brain become necessary for survival? From a moral standpoint it’s going to be very difficult to construct a framework that not only works today, but evolves over time so meet the ever-changing views of society. Case-in-point, the laws of society have changed over the centuries and if we do develop a future moral and legal framework for computers to abide by, it must be dynamic.
Users who have LIKED this comment:
Thank you for your interesting post Victor!
Most of us usually focus on the growth and development of AI itself, but the computerization of humans are not discussed as much. I believe that this is an area carrying great importance in terms of the technologies of the future and also ethical issues. I did not have much knowledge about Neuralink prior to reading your post. I found the project to be very intriguing. It reminded me of the movie “Transcendence” where they upload a man’s consciousness to a quantum computer, and I would recommend it as it made me think of this topic from another perspective. One other important point you raised was the possibility of the non-biological parts of a synthesis like this, dominating the biological parts. I was curious about your further opinion on this issue. Having robots take over the world is not realistic but do you think most people will have some kind of technology implemented in them? Such as having chips for healthcare or even connected brains to computers?
Users who have LIKED this comment:
Thank you for your comment Romi.
From an empirical perspective, we’ve seen that futuristic prediction, especially when new technologies are involved, usually end up being completely wrong.
Sci-Fi movies are actually a good example of it. We’re in 2017 and I don’t see any flying cars around…
As for the “robot revolt” theory, or what I would call the AI takeover, I actually consider it quite realistic. In the long-term, it will be a 50-50 situation: either we manage to harness the power of AI and evolve as a species or we let it end us.
Hello Victor,
I enjoyed reading your post mainly because it demonstrates that the complete blur in the cognitive processes of humans and computers will occur much sooner than many would expect. Building off on your note that visual and speech recognition among computers have either reached or transcended those abilities in humans, I was surprised to discover a biometric, facial recognition feature on a laptop I bought very recently. While I knew modern AI was capable of this rather simple task, many technologies such as this have accelerated particularly in the recent years. Ten years ago, many would have found it difficult to believe that AI could be capable of forming accurate responses to rather vague vocal commands such as Alexa does today. While the computer has not crossed a level of abstraction to provide it with a degree of independence and character– which would be a very large hurdle to tackle indeed– recent booms in AI could mean a paradigm shift of how humans not only view machines but themselves as well. As you have pointed out, being able to computerize the human brain means that our entire state of mind is in our grasps– an idea which is frankly quite unnerving. It will be interesting to observe the process of the human brain being seen from a non-biological perspective.
Sorry for the extra bit here. I’m part of the 238 A class.