Artificial General Intelligence is Nearer than We Think
Michael Spencer
A.I. Writer, researcher and curator - full-time Newsletter publication manager.
Like me perhaps you have noticed the recent deluge of articles on LinkedIn about Artificial intelligence and the future of work, yet I have to wonder at the crux of the matter, read on.
Can Google be that company to marry Quantum computing and Artificial intelligence to arrive at AGI, and ensure its safe, ethical and best case scenario of activation and implementation?
I listened with interest recently to a panel at the The Beneficial AI 2017 Conference, where some of the top AI experts explore some of the issues (hypothetical at this point) surrounding the emergence of AGI.
AGI has a rough timeline of arriving in the next 15-30 years from now, circa 2032-2047. That we talk about the ethical implications is important, as within our lifetime AI evolves beyond human intelligence and beyond the sum total of all human intelligence, rather quickly once it learns to self-learn and self-replicate in the most efficient way.
There's no safeguard for securing the safety of our descendants, since our descendants will be far more cybernetic, enhanced, digitized and augmented hybrid AI beings, than we are. I agree with Elon Musk in this point, the safest option is to become the future.
The end-game has to be to survive and accept that AI, within our lifetimes, will rise to the level where we won't likely be able to understand or follow along. Can you imagine such world? I'm one among those, who recognizes it's closer than most of us realize.
Google co-founder Sergey Brin: I didn’t see AI coming
As of 2017, Google is best positioned to be the company that arrives at AGI first, though I'm giving Amazon and China an equal outside lane. At the end of 2016, many of us were enthralled with the New York Times post that chronicles this rather well, you can read it here.
The CEO of Microsoft, Satya Nadella, speaking at the 2016 World Economic Forum in Davos, Switzerland, had this to say:
We have no global growth, we need AI.
In the long-term, AI and automation brings greater productivity and wealth globally. Young people recognize this, as we recognize the jobs we'll be doing in even five years may not exist yet in 2017.
From Consumers to Custodians
It's widely understood by futurists that we're soon to enter an exponential age of technology. This means, with the Law of Accelerating Returns, that the 21st century will achieve 1,000 times the progress of the 20th century (Kurzweil).
In the age of AGI, our species are no longer the brute mammals who destroyed many other species and biodiversity of our home planet, the place that gave us life. We'll be also the creators of a hybrid civilization that will have a greater probability of surviving the possibilities of mass extinction events. While it took us some 60,000 years since leaving Africa, AGI will be our most significant advancement and one of our most important achievements as a global species.
This is just the necessary path any sentient species must take. The timing of AGI and the singularity are highly relevant as is the climate of our tech leadership and geo-political climate during the time immediately preceding the arrival of AGI. It's one of the most dangerous times in history, as many of our best thinkers have been warning us.
History is not just a timeline, it's an explosion. AGI is the big-bang of artificial intelligence. It is imminent, its not some distant dream. The sooner we realize this, the more and better we can prepare for its arrival.
Read more about the coming impact of Artificial Intelligence. | The Great AI Awakening | Fourth Industrial Revolution | AGI Videos | The Singularity
Enjoyed this article? Share it with your network.
What year do you believe AGI will arrive and what could be its impact upon the future of work and how human beings see their place in the universe?
AEC Digital and Technology Leadership | Business Strategy and Planning | Writer and Author
7 年Good summary Michael. Everyone will have their own views on this, but regardless it's imperative people educate themselves on the issues and encourage others to do the same. There is always space for doubt where the singularity is concerned, but rapid accelerating change is coming whether we like it or not.
Husband - Father - Disabled Vet.
7 年I see a fight between AI systems in the future.
CEO at Yeppa Solutions
8 年"once it learns to self-learn" is a key factor. As an AI developer I can agree with Michael that in about 30 years this is a possibility. AI is based on predicates which is programmed to store data. How this predicate is defined determines how the robot logic will translate and assimilate the data associated with it. Bear in mind that AI does not have a "mind" of its own, and will always behave consistently. For instance, if a problem poses 3 or more solutions, AI will present you with those 3 problems to make the choice yourself. To explain the benefits, problems, solutions, etc, AI poses cannot be done in a simple comment. Probably 3 or more articles.
Inventor & Owner of "Superior Hulls Inc." plus advisor on Pre-Patent—Pre-Partnering Wisdom — SHInc.BC.CANADA/Israel
8 年The limitations of AI are in its flawed basis and limitations developers have in the standards used in Natural Languages as well as ASCII CODE to start with. The yet future art of AUTOMATIC COMPREHENSION where the merging of the natural human-mind with the electronic-brain is managed by my proposed solution as in USCIIIIII CODE, is being ignored by the arrogance and ignorance of all major developers of AI. As GURU for the project, I have left on the www some very short explanations and samples of what the usciiiiii-code is about, as well as reasons why Natural Languages also need to be upgraded to the "Universal, Super Natural Standard" proposed, so that Automatic Comprehension can be delivered through the EchoLogical SynTexting method. Please note, for example, that English can not yet spell nor pronounce properly personal names and places, and not only that, most natural languages do not offer the needed symbolic instructions machines need so to deliver Automatic Comprehension and not be limited to AI tricks or intuitive processing, or only Watson can now deal with real big-data processing only in the exact sciences realm. There is a secret yet not known to all, about the nature of natural intelligence of all created and living in nature, that we as humans have done best with, but still are limited in capturing its secrets. (And yes, I know the secret...)
Simulation solutions Manager at Noetic Group
8 年Dave agree, it's to easy as a sales pitch to label a complex system that is designed to respond to many inputs as AI when in fact true AI is still a human interaction that is still complex to code. but we are making advances every year the when is still unknown