Doug Rose’s Post

Artificial intelligence (AI) has transformed the ancient game of Go, a strategic board game originating in China over 2,500 years ago. AlphaGo, an AI developed by DeepMind, achieved a remarkable feat by mastering Go, a game exponentially more complex than chess, with more possible positions than there are atoms in the universe. AlphaGo’s victory over the reigning world champion in 2016 showcased the extraordinary potential of AI to excel in tasks once deemed too intricate for machines, highlighting AI’s growing capability to achieve superhuman mastery in complex fields. AI has since made major inroads in all fields of study and has gotten the attention of many people who believe that humanity will evolve significantly with the evolution of AI, machine learning (ML) and robotics. #AI #ArtificialIntelligence #Go #AlphaGo #Innovation #Technology

My friend and mentor, Doug Lenat, who found ways for us to continue working on General AI even when it was very unpopular, died a few days ago. And far too early. When AI systems are able to help us with the hard parts of thinking, vastly accelerating scientific progress and the advancement of human civilisation, it will be in part because of Doug's dogged determination to solve hard problems and to understand how human reasoning works. Fundamental ideas behind Cyc: heuristic formal reasoning, choosing appropriate levels of formalisation. forward exploration of entailments to limit the effects of inappropriate formal entailments, structured relationships between theories with contextual applicability -- fundamental ideas in knowledge representation and reasoning that Doug pioneered -- will strongly influence our path towards AI that expands human capability and and wellbeing. Doug's influence, and the influence of the people who joined him in building Cyc, will endure for centuries. Thank you Cade Metz for this very nice overview in the NYT of Douglas Lenat's life and works, and to Gary Marcus for helping to make sure it happened. And thank you especially to Mary Shepherd, R. Guha, Keith Goolsbey, Dave Schneider, Stuart Russell, Ben Rode, David Gunning, Amanda Vizedom, Cynthia Matuszek, Stephen Garrison, Abhishek Sharma, Karen Pittman, Diana Puglisi, Robert C. Kahlert, Zelal Gungordu, Larry Lefkowitz, Liz Coppock, Michael Pool, James (Chip) Masters, Preetha Appan, Ronald Loui, John Cabral, Doug Foxvog, Jared Friedman, Max Crouse, Stephen Reed, Mike Reimers, Michael E. Stewart, Todd Hughes, William Jarrold, Marko Grobelnik, Blaz Fortuna, Luka Bradesko, Dunja Mladenic, Mitja Jermol and so many, many others who have helped make AI more real and more capable. It is often very tempting to put scientific approaches in opposition. At the moment, some would have us take sides in a "logic (or GOFAI) vs LLM" divide. Doug, while doing what was necessary to sustain the Cyc project, was always at heart a true scientist, acknowledging the potential of neural networks in concept definition more than a decade ago, enthusiastically supporting early experiments in reinforcement learning for theorem proving, and, in our most recent discussion, a few months ago, embracing the potential for Cyc to have a role in training learning-based AI to be better at reasoning, and for LLMs to contribute to building up Cyc. He often talked about the Cyc project "priming the pump for AI"; it would have been wonderful if he could have lived to see that dream fully realised, but at least he lived to see it start. Perhaps Peter Beck, you might be able to help Mary Shepherd with the project of spreading some of Doug's ashes on the moon. He certainly aimed for it.

In the early 2000s, Marvin Minsky—a cognitive scientist and one of the fathers of artificial intelligence—liked to say that we can make a computer capable of beating the reigning genius of chess, but we can’t make a robot capable of walking across the street as well as any normal two-year-old child. The real world is not a strictly regulated closed system like a chess game. Sensing moves on a virtual board and responding within agreed-upon rules is one thing. Sensing and physically responding in actual reality—where a huge number and type of unexpected events might occur—is quite another. In fact, the entire AI industry has been through multiple attempts since the 1950s to grow into a mature market. Each of these efforts collapsed because the technology was unable to meet the unrealistic public and investor expectations generated by non-real-world computing triumphs like those of IBM’s Watson. https://zurl.co/IHnF

The AI/ML is overkill and a stupid thing when the problem is well-known mathematically, like the dynamic traces of routing and connections between drivers and customers in #Uber. These Uber guys use a very expensive AI/ML platform with zillions of data in a distributed database. In reality, mathematically, you can reduce the problem to a multiple domination graph problem. Yes, you can argue that it is an NP problem and you need the brute force of zillions of AI mules. However, modern graph theory combines set theory with combinatorics and probability theory, allowing you to transform an NP problem into a P problem by calculating upper and lower bounds. So, in the end, you only need a simple, cheap laptop for that, given the multi-domination theorems and a very simple algorithm with quadratic O complexity, making it a P-solvable problem instead of the NP brute force of Uber.

𝐋𝐋𝐌𝐬 + 𝐔𝐧𝐢𝐭𝐲 + 𝐕𝐨𝐢𝐜𝐞 = 𝐂𝐨𝐧𝐯𝐞𝐫𝐬𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐄𝐱𝐩𝐞𝐫𝐢𝐞𝐧𝐜𝐞 In this video a group of AI is trying to figure out who among them is a human. First they ask each other a general question and then evaluate the answers with some reasoning on why the following character is or isn’t human. To create such experience different LLM providers were used: OpenAI, Anthropic, Google, Meta and video game engine Unity. The experience is enhanced with realistic sounding voice powered by ElevenLabs Each character or NPC (non-player character) has their own SYSTEM prompt describing their behaviour. Intro and outro are scripted. Watch for yourself and let me know what you think! Created by Tore Knabe #conversationalai #generativeai #chatgpt #innovation

🌟 Remembering Douglas Lenat: A Pioneer in Artificial Intelligence 🌟 It is with great sadness that we announce the passing of Douglas Lenat, a brilliant mind in the field of artificial intelligence (AI). He dedicated nearly four decades of his life to unlocking the potential of AI and imparting common sense to machines. As a professor at Stanford University, Douglas developed Eurisko, an AI system designed to discover new scientific concepts. In an astonishing display of intelligence, Eurisko analyzed the complex game Traveller Trillion Credit Squadron and identified winning strategies. Armed with its unconventional approach, Douglas triumphed in a tournament and even won again the following year after the rules were changed. Inspired by this success, Douglas embarked on the Cyc project. This ambitious endeavor aimed to define the fundamental laws governing our world, using a rule-based approach to replicate human common sense. Despite the current focus on data-driven AI in the tech industry, the Cyc project embodied Douglas's unwavering belief in the power of handcrafted intelligence. Throughout his life, Douglas was known for pushing boundaries and never shying away from challenges. He achieved remarkable feats in the pursuit of AI's potential, even as the field faced periods of stagnation. His work on Cyc spanned over 2,000 human years and produced more than 25 million rules. Beyond his contributions to AI, Douglas was an adventurer who explored over 100 countries and all seven continents. His legacy will continue to inspire future generations of AI researchers. Join us in honoring Douglas Lenat's incredible journey and the impact he made on the world of artificial intelligence. His unwavering determination and groundbreaking work will forever shape the future of AI. Rest in peace, Douglas Lenat. 🙏 #ArtificialIntelligence #AI #InMemoryOf #Technology

You Cannot Be At Two Places At Once = You say : Duh! It is obvious. Please understand this: ChatGPT is not rule based - no rule, however obvious to us, is known to ChatGPT. Case in point is a recent experiment by Pranab Ghosh on a simple Blocks World problem : I have A on top of B on top of C - get me C ==> B ==> A (or something similar). While solving this - ChatGPT violated rules of the Blocks World. In reality it means simply this: ChatGPT intelligence (if it has any intelligence) will not be suitable for Robotics. Robots need to follow rules like not bumping into humans or trample babies. = Shekhar Veera let me know about the Cyc Project today. Cyc is a long-term artificial intelligence project that aims to assemble a comprehensive ontology and knowledge base that spans the basic concepts and rules about how the world works. Hoping to capture common sense knowledge, Cyc focuses on implicit knowledge that other AI platforms may take for granted. This is contrasted with facts one might find somewhere on the internet or retrieve via a search engine or Wikipedia. Cyc enables semantic reasoners to perform human-like reasoning and be less "brittle" when confronted with novel situations. Douglas Lenat began the project in July 1984 at MCC, where he was Principal Scientist 1984–1994, and then, since January 1995, has been under active development by the Cycorp company, where he was the CEO. The Cyc project has been described as "one of the most controversial endeavors of the artificial intelligence history". Catherine Havasi, CEO of Luminoso, says that Cyc is the predecessor project to IBM's Watson. Machine-learning scientist Pedro Domingos refers to the project as a "catastrophic failure" for several reasons, including the unending amount of data required to produce any viable results and the inability for Cyc to evolve on its own. Gary Marcus, a professor of psychology and neural science at New York University and the cofounder of an AI company called Geometric Intelligence, says "it represents an approach that is very different from all the deep-learning stuff that has been in the news.”[30] This is consistent with Doug Lenat's position that "Sometimes the veneer of intelligence is not enough". Every few years since it began publishing (1993), there is a new Wired Magazine article about Cyc, some positive and some negative (including one issue which contained one of each). = I will stop here. Now we are at a juncture; We have billion dollar LLMs which do not know about basic obvious rules like you cannot move a block on which another block is resting. On the other hand promising projects like Cyc will not see the light of the day since it will not get the billions of dollars it may need. What is the conclusion: People like Pranab Ghosh should stop expecting intelligence out of AI. In the following I have suggested 3 definitions of AI: 1) AI : Aspirational Intelligence 2) AI : Artificial Ignorance 3) AI : Absurd Intentions

Generative Adversarial Network, an exciting field, which is a magical wand in the tech world these days, be it for generating Fake human faces, transition of a sunset image into a beautiful sunrise image, text to image translation, turning aged faces into young faces, and what not. 🪄 To understand the logic behind this magic, it's very important to understand the game of Two Players. Player 1: 🤹 The name of player 1 is Generator, who is responsible for generating Fake content, the generator wins, if it's generated fake content is proved as original by Player 2. Player 2:🤹 The name of player 2 is Discriminator. It is responsible to discriminate between the original and the fake content. It is the core aim of the generator to fool the discriminator into believing that the samples generated by it are real. And this duo of the generator and discriminator, is the basic architecture of GAN model.✨ Follow for more such technical content🖥️ #gans #artificialintelligence #neuralnetworks #deeplearningai #machinelearning

I became exhausted managing an ever-expanding list of GPT Prompts. After exploring multiple tools without success, I turned to AI to develop a user-friendly tool. This tool not only stores prompts but also allows for prompt sharing and provides access to shared prompts from others. If you're interested in trying a tool that has been almost completley generated by AI, feel free to give it a go, check it out at https://lnkd.in/gjVkYupy

Think of the human brain as a ‘90s computer trying to run the latest 4K video game. Sounds like a good idea, right? Well, welcome to the world of artificial intelligence, where a new breakthrough pops daily, brighter than New Year’s fireworks. Our brain, armed with good intentions and a coffee maker, decides to get a grip on it. Before you know it, it’s navigating through a sea of AI promising to make life easier. From an app that sweet-talks you more than your cat to another promising to turn you into the digital era’s Ernest Hemingway, our hero tries to keep up with everything. The result? Smoke. Literally. The cerebral CPU hits record-breaking temperatures, so high you could cook an egg on it. As synapses start sending SOS signals and the head turns into a disco with strobe lights and alarm sirens, the epiphany hits: “Wasn’t all this supposed to make life easier?” Well, in a way, yes. But no one said learning to do so would be a piece of cake. And so, friends, here we are, with a brain looking more like a summer BBQ than a thinking organ. But hey, it’s all grist for the mill (or should we say “all grist for the code”?), because in the end, this is the beauty of learning. And yes, it might take a bit to figure out how not to turn your brain into an omelette trying to keep up with AI, but once you’ve got the knack, it’s all downhill from there. Moral of the story? The learning curve might seem more like a roller coaster with endless loops, but once you learn to enjoy the ride (and maybe use a fan for your brain), you discover that in the end… well, it’s all a big laugh! A heartfelt shoutout and thanks to the countless many — I can't possibly name you all, and let's be real, some of you don't even know I exist. Your relentless passion and mind-blowing creativity have inspired me and kept my fire (and focus, in the photographic sense) alive and kicking. Here's to you all, unknowingly fueling my drive and sharpening my lens on the world. Cheers! Brian Sykes, Christopher Sicurella ∴, Margarida Barreto, Maribeth Woodford, 🚀 Robin Doris,Seth Baum, ai James Larkin, Sandu Baciu 👨🚀, Alexander Esau, Dogan Ural John Chao 🧧, Dan Frydman, Hugo Barbera 👁️👁️👁️, Abhinav G., Ken Newton, Big Al Gruswitz, José Parra (Chema Parsanz), Tianyu Xu,Nejc Susec, John Vander Weit III … #ArtificialIntelligence #MachineLearning #AIimagination #AIart #AIStrategy #AIevolution #Innovation #TechTrends #DigitalTransformation #FutureOfWork #AIEthics #AIResearch