Pathways, Timelines and Superintelligence Scenario

Wait but Why summarizes some of the timelines and definitions for artificial super-intelligence. The general consensus among AI researchers is that Artificial superintelligence could arrive around 2060.

Nick Bostrom defines superintelligence as “an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills”.

Less Wrong has some visualization and summary of superintelligence pathways

This displays the five pathways toward superintelligence that Bostrom describes in chapter 2 and returns to in chapter 14 of the text. According to Bostrom, brain-computer interfaces are unlikely to yield superintelligence. Biological cognition, i.e., the enhancement of human intelligence, may yield a weak form of superintelligence on its own. Additionally, improvements to biological cognition could feed back into driving the progress of artificial intelligence or whole brain emulation. The arrows from networks and organizations likewise indicate technologies feeding back into AI and whole brain emulation development.

Artificial intelligence and whole brain emulation are two pathways that can lead to fully realized superintelligence. Note that neuromorphic is listed under artificial intelligence, but an arrow connects from whole brain emulation to neuromorphic. In chapter 14, Bostrom suggests that neuromorphic is a potential outcome of incomplete or improper whole brain emulation. Synthetic AI includes all the approaches to AI that are not neuromorphic; other terms that have been used are algorithmic or de novo AI.

NBF believes that the primarily hardware based approaches to superintelligence tend to be more narrow solvers. If very good quantum computer computers are realized. The systems are cracking mathematically hard problems and assisting the process to improve machine learning.

Powerful solvers have human involvement to put in and get out answers and do not have the runaway AI scenario.

Before 2060 there should be
There should be advanced Optalysis style optical computers that have implemented Deep learning.

A startup company called Optalysis is trying to invent a fully-optical computer that would be aimed at many of the same tasks for which GPUs are currently used. Amazingly, Optalysis is claiming that they can create an optical solver supercomputer astonishing 17 exaFLOPS machine by 2020.

Deep Learning + 17 exaFLOP optical computer = 17 ExaFLOP Deep learning system by 2020.

The GPGPUs that implemented the Baidu Deep learning brain may be replaced by new optical computers.

Deep learning is a hot AI (artificial intelligence) field now. It is being used to develop AI that can perform tasks like learning to play video games better than humans

There should be quantum computers with trillions of qubits. Currently DWave has a system with about 2000 qubits and they are doubling the qubits every year. There is work towards creating qubit using quantum dots. There is work towards using quantum dots to create general purpose computing using cellular automata.

Inside of Dwave Systems Quantum Computer

Advanced wearable systems are providing people with easy access to the internet and google search and wikipedia and youtube for optimized access to existing human knowledge.

Memristors should make the largest scale neuromorphic systems

Memristors can mimick the electrical properties of synapses. HP is betting the company on making universal memory with memristors and optical onchip networking. Within 20 years this should lead to multi-ExaFLOP supercomputers with trillions of neuromorphic synapses.

Human genetic other enhancements

There is progress to intelligence enhanced humans via genetic engineering

* about 16% of people have trouble conceiving, older couples have more probems and need invitro fertilization
* 1.5% in US currently use invitro fertilization (IVR)
* China will lift the one child ban completely within 5 years
* IVR is getting cheaper ($20K to $1K or less) and more effective for each treatment (32% to 60% or more)
* IVR makes it easier to choose genetic engineering (embryo is already out)
* Chinese and Asian cultures are more enthusiastic about making the children better through any means

Lower costs, more effective IVR and better CRISPR genome engineering should see IVR/genetic engineering heading towards 2-10% of population.

This can lead to the scenario of millions of “Tony Stark” like supergeniuses before any true synthetic superintelligence. The genetic and pharma enhancements would also be able to provide safe steroid effects and bones that can break surgical drills. Some people have the genes for very strong bones. You could also dial up the speed of healing and regeneration.

Ido Bachelet’s DNA nanobots mean that adults could also be genome modified in a couple of decades

At the British Friends of Bar-Ilan University’s event in Otto Uomo October 2014 Professor Ido Bachelet announced the beginning of the human treatment with nanomedicine. He indicates DNA nanobots can currently identify cells in humans with 12 different types of cancer tumors.

A human patient with late stage leukemia will be given DNA nanobot treatment. Without the DNA nanobot treatment the patient would be expected to die in the summer of 2015. Based upon animal trials they expect to remove the cancer within one month.

Within 1 or 2 years they hope to have spinal cord repair working in animals and then shortly thereafter in humans. This is working in tissue cultures.

Previously Ido Bachelet and Shawn Douglas have published work on DNA nanobots in the journal Nature and other respected science publications.

One Trillion 50 nanometer nanobots in a syringe will be injected into people to perform cellular surgery.

The DNA nanobots have been tuned to not cause an immune response.
They have been adjusted for different kinds of medical procedures. Procedures can be quick or ones that last many days.

Medicine or treatment released based upon molecular sensing – Only targeted cells are treated

So by 2045

There should be

* beyond trillion qubit scale quantum computing
* advanced DNA nanotech and nanobot surgery
* other molecular nanotechnology
* beyond trillion synapse scale neuromorphic computing
* some level of whole brain emulation
* Zettaflop computing or beyond
* advanced optical computing
* significant genome modification


None of this looks like the fast emerging or secretly created singleton scenarios.
It looks like a lot of slow hard work and broad progress to a civilization with multipolar superintelligence.
Millions to a billion augmented human minds utilizing narrow AI, quantum computer and optical computer solvers.

SOURCES – Nick Bostrom, Less Wong, Youtube, Waitbutwhy