# Paths To Truly General AI By [Madhav Goyal](https://paragraph.com/@madhavg) · 2021-12-18 --- AI is one of the most common buzzwords in tech today but the AI being used today in every device is not the true vision of the pioneers who introduced this concept. Some of the greatest Computer Pioneers asked these questions if ever a machine could think like us or ‘imitate us’ as Alan Turing put forward in his [paper COMPUTING MACHINERY AND INTELLIGENCE](https://watermark.silverchair.com/lix-236-433.pdf?token=AQECAHi208BE49Ooan9kkhW_Ercy7Dm3ZL_9Cf3qfKAc485ysgAAAtEwggLNBgkqhkiG9w0BBwagggK-MIICugIBADCCArMGCSqGSIb3DQEHATAeBglghkgBZQMEAS4wEQQMxDNR8cuhxw0ZIbwNAgEQgIIChLz7LftHmNF7YMqwv8RpnHOTgQqLDh1wwJz20XxG4AbNL3pkK0P-_6sleKFkGB7Jo0y8102fjjgVWNxDoTReWUq_CeBAmrTvpGNAo6COr9pCHgihIOweDZ5-oWdhnPzEYtw20VwDd1mjzz0P2eFFoTsrOABZlRAJMZLxdYjHkb9U4xM31wyXJibW2Z9hAJAc2Zg5uQAlrBQvbHZs61Ap1aHLsYyDAym4Km1soyYN6OHGpJ_T2H1JOUnzTx3EnN6HmW_SklhvHOQqshm5521zcPNhb4JcUW0mawUhzKNKKrkdld-hF-7SKF3GSeZ11IKxmCuC9a8Sfw0-HajQBB35WenomQO8MUA8YdacjgY8zOOskbzAKAzTNUc0-as13DBUaNSaToPRIFyvBlZAfk9I8b1hoNI1pbsGAh700RLbtIE0TrMotQrk4asYhegqJfCA9atrZOzhzMJAFjdWI1oVf8LKfTl-rzWtVxYBooImDhXcP1aTKVkH5MHir2RsY0WuxaKnGrFm99IDmBzwwttWfc7jmS9BU18WE4eIdRm1QzHjQy4l-xIJOHxXwu7baTwv4kJjlX0zKEbs6HTiXZvt9ZcqIwi9-jftHje4_TRO5aBzGb8drGImeuSNU5sJY7UUSL2u85d7ub5FxS6ag0AyqdeMWIKmgruu5f4G6U7oiy7LIFxtMxcDwkVyIeq0XGoZ0102uc7vZnc-zUKG9JIHObeQyeLGc44JqKCdpYfBEV_SknIwUWOW90KO45b8EL72sWd8UgacOqqIXirDMv-55a9PLZUFukZDc_niNcBLGQVASMxvzBHafD_0EcpvsuVlsk81V4FVtRq9Fj4tr7G8vxqoYZSn). One of the most fundamental questions we have is “ what is consciousness” which we haven’t quite figured out yet. Finding the answer to this question is one of the root goals of making a brain that could do general-purpose tasks like us as there is probably no better way to understand a topic than creating it. Some of the most beautiful minds of our time spend their time encountering this idea and pushing it towards reality. There has been quite a lot of arguments about which approach towards AI could lead to AGI. Most of them agree that there could be several paths to a thinking machine. The most openly talked about and known is [Deep Learning](https://booksc.org/book/40797728/06cf0c) - whose structure loosely resembles that of our brain. It constitutes interlinked artificial neurons to which we train and teach the specific task through data. It is one of the most revolutionary technologies, it is being used in from Drug discovery to farming practices to Natural language processing. This approach has been the most hyped and at the forefront of the AI revolution this decade. Its origins can also be traced back to Frank Roseblat’s perception on which I did a separate [post](https://madhavgoyal.substack.com/p/perceptron-the-foundation-of-all). The next most popular approach also is Symbolic and capturing common sense which is pursued by Cyc a long-term Ai project which focuses on [implicit knowledge.](https://en.wikipedia.org/wiki/Implicit_knowledge) It tries to describe aspects and things in a certain no. of steps from the fundamental root. When Cyc was founded it was estimated it would take around 1000-3000 person-years and yes it’s still going on as you would have guessed. It’s kinda hard to describe this approach other than outlining 2-3 words so I hope you delve deep into it if you hope will link some resources. Now comes my most favorite approach -the [CogPrime Architecture](https://wiki.opencog.org/w/CogPrime_Overview). It tries to bridge neural networks and symbolic ways hope to achieve AGI with the Cognitive synergy of all of them. the reason I think it’s the most promising is that it’s not a single way but a combination of the best ones. It uses hypergraphs as the knowledge representation way to represent Cogprime’s cognitive processes. Its AI agents use [Atom Space](https://wiki.opencog.org/w/AtomSpace) a knowledge representation database where knowledge flows in the form of Atoms. Most of the stuff of the OpenCog is present in Atomspace. As old as cinema, people have made movies on robots taking over our work and rendering us jobless in misery which is only partly true. I think that when we make or develop an AGI( Artificial general intelligence) it will be one of the most fruitful accomplishments of Humankind. It will open the doors for us which we never even knew existed and help us accomplish the most daunting tasks and answer the most fundamental of questions and that can only happen if our and AGI’s goal aligns with each other which is kinda unknown area but constant progress is being made in every aspect of making a safe future for Humanity. There has been no concrete evidence to this answer as to whether “machines” could ever be conscious or think but there have been a lot of things that seemed unlikely and critiqued which in the end turned out to be true. PS Thanks dear reader for reading the topic that is dearest to my heart . I thought of making the blog long but it might become a hassle to read for others so . The topics I talked about are critiqued a lot but it’s one of the main things that leads to better outcomes. If you want to show a gesture of appreciation and want to keep getting blogposts from me buy me a [book](https://www.buymeacoffee.com/madhavgoyal) https://www.buymeacoffee.com/madhavgoyal or I will just drop my eth address just for fun 0x5c0265DCFBeeA7A363098483fE4910052D4DD668 . [Mint an NFT and be part of the revolution](https://www.getskywalkerz.com/) https://www.getskywalkerz.com/ --- *Originally published on [Madhav Goyal](https://paragraph.com/@madhavg/paths-to-truly-general-ai)*