Our team has been developing data technologies for artificial intelligence (AI) since 1998, two years before we applied for our first patent of a standardized approach for automated Data Integration and Consolidation. Since then, team members have been named inventors of twenty-two United States patents and one current US patent applications.
We are currently filing twenty-five patent applications for technologies that will facilitate the next generation of AI. The next generation will be based on so-called neuro-symbolic AI that combines the use of machine learning technologies utilizing “neural networks” and knowledge representation and reasoning technologies utilizing “mathematical symbols”.
Today, computers are highly proficient at collating very large amounts of information as input and regurgitating it as elegant output.
In November 2022, OpenAI announced ChatGPT, a conversational chatbot utilizing Generative AI, or “GenAI”. Despite the hype, ChatGPT remains like all “chatbot” technologies developed since Joseph Weizenbaum's program “ELIZA”, published in 1966, about which Weizenbaum opined
“In artificial intelligence, machines are made to behave in wondrous ways, often sufficient to dazzle even the most experienced observer. But once a particular program is unmasked, once its inner workings are explained, its magic crumbles away; it stands revealed as a mere collection of procedures. The observer says to himself "I could have written that". With that thought, he moves the program in question from the shelf marked "intelligent", to that reserved for curios. The object of this paper is to cause just such a re-evaluation of the program about to be "explained". Few programs ever needed it more.” [Source]
On 29 August 2023 it was reported, “Chat-GPT is “much smarter” than the average human with an estimated IQ of 155, says former Google executive Mo Gawdat.” [Source] Mr Gawdat is a systems engineer, entrepreneur, author, podcaster, and public speaker and now hosts a popular podcast and has written several books. ChatGPT does not have an IQ of 155.
It's worth noting that in 1970, AI pioneer Marvin Minsky told Life Magazine, “In from three to eight years we will have a machine with the general intelligence of an average human being.” [Source] We are still waiting.
Despite many billions of dollars investment, we still can’t be confident that the output of Generative AI chatbots has integrity, that it is factually accurate. The legendary tales of ChatGPT producing misinformation is testament to that. We say that if a system cannot recognise misinformation, then it is not intelligent.
We say that AI is two to five years away.
IBM is the world’s longest standing entity for AI research since its participation in the 1956 Dartmouth Summer Research Project on Artificial Intelligence. No other company in the world today has invested so much time in AI research and development over so many years.
Around 2016, IBM referred to its “Watson” AI platform as a cognitive computing system. [Source]
In May 2023, six months after OpenAI announced ChatGPT, IBM announced “Watsonx”, a “portfolio of AI products that accelerates the impact of generative AI in core workflows to drive productivity”. [Source]
Despite research into hybrid AI (the combination of machine learning and knowledge engineering) dating back to the 1980’s and symbolic AI (the use of mathematical symbols) dating back to the 1990’s, it wasn’t until 2024 that IBM opined publically that “neuro-symbolic AI was a likely path to a generation of AI beyond the current GenAI, saying “We see Neuro-symbolic AI as a pathway to achieve artificial general intelligence. By augmenting and combining the strengths of statistical AI, like machine learning, with the capabilities of human-like symbolic knowledge and reasoning, we're aiming to create a revolution in AI, rather than an evolution” [Source]
We have arrived at a similar point in our thinking about our cognitive AI platform because we believe that computer cognition requires a combination of many technologies including machine learning with neural networks and knowledge representation and advanced automated reasoning with mathematical symbols.
The next step on the path to true artificial intelligence is the expected advance of computer reasoning techniques and technologies that GenAI is incapable of no matter how many hundreds of billion dollars of high speed computing chips are provided to it.
In 1995 it was said that “Angelo Dalli,[6] Henry Kautz,[7] Francesca Rossi,[8] and Bart Selman[9] also argued for such a synthesis. Their arguments attempt to address the two kinds of thinking, as discussed in Daniel Kahneman's book Thinking, Fast and Slow. It describes cognition as encompassing two components: System 1 is fast, reflexive, intuitive, and unconscious. System 2 is slower, step-by-step, and explicit. System 1 is used for pattern recognition. System 2 handles planning, deduction, and deliberative thinking. In this view, deep learning best handles the first kind of cognition while symbolic reasoning best handles the second kind. Both are needed for a robust, reliable AI that can learn, reason, and interact with humans to accept advice and answer questions. Such dual-process models with explicit references to the two contrasting systems have been worked on since the 1990s, both in AI and in Cognitive Science, by multiple researchers. Without a doubt, the most important, and likely the last major challenge for AI delivery, is to use a mix of known techniques in a new way to develop cognitive AI that will mimic aspects of human reasoning necessary to distinguish fact from fiction”.
However the 1990’s debates about rules-based, connectionist, similarity-based, and case-based reasoning of the time did not yield an answer. It merely demonstrated the difficulty of emulating the mysterious capability and capacity of the human brain. And as at today, it informs the reality that for many years to come, humans-in-the-loop will very much be a mandatory feature of robust and responsible AI.
