Summary:<\/em><\/strong> \u00a0Researchers in Synthetic Neuro Biology are proposing to solve the AGI problem by building a brain in the laboratory.\u00a0 This is not science fiction.\u00a0 They are virtually at the door of this capability.\u00a0 Increasingly these researchers are presenting at major AGI conferences.\u00a0 Their argument is compelling.<\/em><\/p>\n \u00a0<\/p>\n Here\u2019s what I think we know:<\/p>\n But what about artificial general intelligence (AGI)<\/strong> that we all believe will be the end state of these efforts?\u00a0 When will we get AGI that brings fully human capabilities, learns like a human, and can adapt knowledge like a human?<\/p>\n \u00a0<\/p>\n How Far Away is AGI?<\/strong><\/span><\/p>\n When I looked into this two years ago the range of estimates was around 2025 to 2040.\u00a0<\/p>\n With a few years more experience under our belt, here are the estimates given by 7 leading thinkers and investors in AI (including Ben Goertzel and Steve Jurvetson) at a 2017 conference on Machine Learning at the University of Toronto when asked \u2018How far away is AGI\u2019.\u00a0<\/p>\n There\u2019s significant disagreement but the median is 23 years (2040) with half the group thinking longer.\u00a0 Sounds like we\u2019re learning that this may be harder than we think.<\/p>\n \u00a0<\/p>\n What\u2019s the Most Likely Path to AGI?<\/strong><\/span><\/p>\n Folks who think about this say that everything we\u2019ve got today in DNNs and reinforcement learning is by definition \u2018weak\u2019 AI.\u00a0 That is it mimics some elements of human cognition but doesn\u2019t achieve it in the same way humans do.<\/p>\n This may or may not also be true of next gen Spiking neural nets<\/u><\/em><\/a> that have adapted some new elements from neuroscience research.\u00a0 They look like they\u2019re a step in the right direction but we really don\u2019t know yet.<\/p>\n There is general agreement that weak AI, while commercially valuable will never give us AGI.\u00a0 Only if we create broad and strong AI systems that mimic human reasoning can we ever achieve AGI.<\/p>\n \u00a0<\/p>\n Is this all on the Same Incremental Pathway?<\/strong><\/span><\/p>\n So far there have been two primary schools of thought.\u00a0<\/p>\n The Top Down<\/strong> school is an extension of our current incremental engineering approach.\u00a0 Basically it says that once the sum of all these engineering problems is resolved the resulting capabilities will in fact be AGI.\u00a0<\/p>\n Those who disagree however say that truly human-like intelligence can never be the result of simply adding up a group of specific algorithms.\u00a0 Human intelligence could never be reduced to the sum of mathematical parts and neither can AGI.<\/p>\n The Bottom Up<\/strong> school is the realm of researchers who propose to build a silicon analogue of the entire human brain.\u00a0 They propose to build an all-purpose generalized platform based on an exact simulation of human brain function.\u00a0 Once it\u2019s available it will immediately be able to do everything our current piecemeal approach has accomplished and much more.<\/p>\n Personally, my bet is on the Bottom Up school though I think we are learning valuable hardware and software lessons along the way from our pragmatic DNN approach.<\/p>\n \u00a0<\/p>\n How About a Radical New Path<\/strong><\/span><\/p>\n What I discovered in revisiting all this is that my own thinking has been too constrained.\u00a0 For example, in writing about 3rd<\/sup> gen spiking neural nets or the neuromorphic modeling approach of Jeff Hawkins<\/u><\/em><\/a> at Numenta I assumed that hardware and software modeling of individual neurons interacting was the agreed approach.\u00a0<\/p>\n Not only is this not true (I\u2019ll write more about this later), but our fundamental assumption about working in silicon is not the only approach being explored.<\/p>\n Folks like neuroscientist George Church at Harvard are proposing that we simply build a brain in the biology lab and train it to do what we want it to.<\/p>\n \u00a0<\/p>\n Computational Synthetic Biology (CSB) \u2013 Wetware<\/strong><\/span><\/p>\n Computational Synthetic Biology (aka synthetic neuro biology) is much further along than you think, and in terms of a 25 year forward timeline might just be the first horse to the finish line of AGI.<\/p>\n As far back as 10 years ago, the field of Systems Biology sought to reduce molecular and atomic level cellular activities to \u2018bio-bricks\u2019 that could be strung together with different \u2018operators\u2019 to achieve an understanding of how these processes worked.\u00a0<\/p>\n If Systems Biology is about understanding nature as it is, Computational Synthetic Biology takes the next step to understand nature as it could be.<\/p>\n Here\u2019s a graphic of the Infobiotics Workbench<\/u><\/em><\/a> from 2010.\u00a0 Anyone familiar with predictive modeling will immediately recognize the similarities with a data analytics dashboard, including setting hyperparameters, loss functions, and comparing champion models.\u00a0 Fast forward to the 2018 O\u2019Reilly Artificial Intelligence Conference in New York where Harvard researcher George Church tells us just how much further we\u2019ve come.\u00a0 (See his original presentation here<\/u><\/em><\/a>. Graphics that follow are from that presentation.<\/em>)<\/p>\n Synthetic neuro biology (SNB) is already ahead of silicon simulations in both energy (much lower) and computational efficiency.\u00a0 Capabilities are increasing exponentially faster than Moore\u2019s Law, in some years by a factor of 10X.<\/p>\n \u00a0 In the upper right, SNB is already operating close to the biological limit of compute efficiency.\u00a0 Church says silicon could catch up in 3 decades at the current growth rate, except that silicon is already plateauing.<\/p>\n Church says \u201cWe are well on our way to reproducing every kind of structure in the brain\u201d.<\/p>\n In the lab, he can already create all types of neurons to order including being able to build significant human cerebral cortex structures complete with supporting vasculature.\u00a0 This includes myelin wrapping of the axons so that signals can be sent over long distances at high speeds allowing action potentials to jump from node to node without signal dissipation.<\/p>\n In short, the position of SNB researchers is it\u2019s easier to copy an unknown (brain function, human cognition, AGI) in a made-to-order biological brain than it is to translate that simulation into silicon.\u00a0 In the silicon simulation you\u2019ll never really know if it\u2019s actually right.<\/p>\n To add one more level to these futuristic projections, what is the possibility that we can modify or augment our current physiology to create super intelligence?\u00a0 The field of biological augmentation is already well underway principally in the field of curing disease.\u00a0 Why not extend it?<\/p>\n For example, the average human brain weighs 1.3 kg and consumes 20W of power (compared to the 85,000W required by Watson to win Jeopardy).\u00a0 What if we could enlarge it to 25kg, about 3X the size of a whale\u2019s brain, but still within the physical limit of what our skeletons could support?\u00a0 It seems the only penalty would be to have to provide about 100W of power, equivalent to eating about 8,000 calories per day.\u00a0 Donuts \u2013 yumm.\u00a0 Could Silicon Based AGI Ever Be Completely Human-Like?<\/strong><\/span><\/p>\n While silicon AGI modelers still struggle to understand brain function sufficiently to build their simulations there are legitimate questions about whether their best result will ever be enough.\u00a0 Will they ever have the capabilities that make us philosophically human?\u00a0 Our science fiction robots have some or all of these qualities:<\/p>\n Consciousness:<\/strong>\u00a0 To have subjective experience and thought.<\/p>\n Self-awareness:<\/strong>\u00a0 To be aware of oneself as a separate individual, especially to be aware of one\u2019s own thoughts and uniqueness.<\/p>\n Sentience:<\/strong>\u00a0 The ability to feel perceptions or emotions subjectively.<\/p>\n Sapience:<\/strong> The capacity for wisdom.<\/p>\n If we construct an intelligent \u2018brain\u2019 in the lab from the same human biological components \u2013 is it still just a simulation?\u00a0 Will it be a \u2018mind\u2019?<\/p>\n So far computational synthetic biological research remains in the lab or where it has matured, is being applied to cure disease.\u00a0 But increasingly researchers like George Church are presenting at major AGI conferences.\u00a0 They are already able to skip the entire step of creating a silicon analog of the brain where most silicon AGI researchers are currently stuck.\u00a0 If there\u2019s a 15 or 25 year runway to develop AGI, I wouldn\u2019t bet against this wetware approach.<\/p>\n \u00a0<\/p>\n \u00a0<\/p>\n Other articles on AGI:<\/strong><\/p>\n In Search of Artificial General Intelligence (AGI)<\/u><\/em><\/a> (2017)<\/u><\/em><\/p>\n Artificial General Intelligence \u2013 The Holy Grail of AI<\/u><\/em><\/a> (2016)<\/u><\/em><\/p>\n Other articles by Bill Vorhies.<\/u><\/em><\/a><\/p>\n \u00a0<\/p>\n About the author:\u00a0 Bill Vorhies is Editorial Director for Data Science Central and has practiced as a data scientist since 2001.\u00a0 He can be reached at:<\/p>\n Bill@DataScienceCentral.com<\/a><\/p>\n<\/div>\n Go to Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"![](\"http:\/\/api.ning.com\/files\/pWFsdFRaG2KjElTlrzqQjUj*YldMOjfafjqxJXPvFbOPfs*t-RVHZGm*t620u-av1Us1SKNa4slHMCodfl0cuUOeFAcfYOZH\/AGIxsmall.jpg\")
<\/a>If you step outside of all the noise around AI and the hundreds or even thousands of startups trying to add AI to your car, house, city, toaster, or dog you can start trying to figure out where all this is going.<\/p>\n\n
\n
<\/a>Source: 2010 GECCO Conference Tutorial on Synthetic Biology\u00a0<\/span> by Natalio Krasnogor, University of Nottingham.<\/span><\/p>\n<\/a><\/p>\n<\/a><\/p>\n