AI Ethicist Views
Where will we be in five years as a result of smart generative AI?
Must of our technology innovations come from the Media and Entertainment industry where AI discards the shackles of human control and takes on a life of its own, as cyborgs powered by artificial intelligence did in “Terminator 2.”
Physicist Stephen Hawking, who died in 2018, described the rise of AI as one of the greatest threats to the future of humanity. Even the controversial Elon Musk and over 1000 technology leaders have called for a pause on generative AI and for advancing legislation.
Francois Chollet, AI researcher at Google is one of the most grounded ethicists in downplaying the impact that large language models (LLM) like ChatGPT, or Google’s Bard chat bots will have on the market. So many people are thinking that this type of technology will create 100X returns or create a “Society and citizen alternative reality or thing, “ he said.
I don’t think I have met one AI Ethicist that is not gravely concerned about the risks inherent in AI, without a vision of the world we want to live in and more clarity of a unified legal framework, and laws put into effective to avoid creating a world where machine is more dominant that mankind.
Sam Altman, founder of OpenView, Chat GPT has also expressed concerns on anyone relying on ChatGPT on anything significant, as it’s still so emergent. Yet the hype has continued.
These technologies essentially are pattern match makers pulling pieces together by absorbing large volumes of data from sources like Reddit, books, news articles, web, and then parsing pieces together to from its rich LLM’s to answer end user questions.
Who would have known that we could generate text, images, sound and other media forms to short questions or prompts. This will bring in a new wave of interactions, where we can have continual two way communications just like we are having between humans.
Venture capitalists are flocking to invest in generative AI companies, and already in 2022 over $140 Billion has been invested. Silicon leading investors at Sequoia Capital predict this type of technology will unleash trillions in value. Even Radical Ventures, a Toronto based VC is in the swing of investing in generative AI technologies, with a new $US500 million fund. Needless to say there is a lot of enthusiasm in chatbots and it’s not just OpenAI’s Chat GPT, also Replika, Character.ai, and You.com are a few of the market leaders to track.
Despite the excitement on generative AI, there are prickly issues to resolve, especially in producing toxic content, surfacing up biases on women, and in particular women of color, copyright infringement, labour market shifts we are not ready to handle are all top of mind issues. The other risk is the hallucination effect of these generative AI technologies creating an answer that is not based on accurate data, (ie: sourced from fake data, fake news) and not having the AI based - large language models (LLM’s) cleanse out erroneous data, creating harm, brand risks, or even impacting stock markets.
AI leader and deep learning expert, Yoshua Bengio, is advocating that AI capabilities have accelerated significantly in the last five years, due to advancements in language-generating technology. In Canada recently, over 75 AI researchers and start-up AI CEO’s have signed a letter advocating that the Canadian Federal Government pass the AI and Data Act, called AIDA, which is part of Bill-C27. The Future Institute of Life also issued in March of this year, an open letter of over 1,000 technology leaders advocating for a 6 month pause to advancing generative AI chat bot technology. The EU has accelerated the plea for the USA, Canada, UK and other democratic countries to rally together to finalize an AI legal framework to create some guardrails around the AI industry.
So what does ChatGPT think of the ethical concerns of AI?
- There are several legal and ethical issues surrounding the use of ChatGPT and other large language models. One legal issue is related to potential copyright infringement, as the model may generate text that is similar or identical to existing copyrighted content.
- Another legal issue is related to potential privacy violations, as the model may generate personal information or sensitive data that could be used to identify or harm individuals. Ethically, there is concern about the potential misuse of the model, such as in the creation of deep fake audio or text or in spreading false or misleading information.
- There is also concern about unintended bias in the training data, which could lead to the model perpetuating harmful stereotypes or discrimination. Another ethical concern is related to the transparency of the model’s decision-making process, as it can be difficult for users to understand how the model arrived at a particular output.
- Furthermore, GPT-3 and other similar models have been trained on a massive amount of data, which may have come from questionable sources, such as scraping of websites or other data collection methods which could have legal and ethical issues.
- In summary, the use of ChatGPT raises legal and ethical issues related to copyright, privacy, misuse, bias, and transparency. It is important for users to be aware of these issues and take steps to mitigate them.
I was impressed by the answer that ChatGPT provided and as this industry matures we must answer more of the challenging ethical questions on AI. I think much of this is rooted in defining what it is to be human and what we want as humans versus having technology leaders continually shape our future. Movies like The Terminator are not impossible to imagine, as movies like The Minority Report are eerily more accurate due to technology advancements of companies like SenseTime, that can monitor walking, traffic, purchasing, and social data into a social ranking profile. Although banned in the USA, the ethical implications of never being able to be unknown or have my private data socially rank my habits has Orwellian overtones.
