Michele M. Merrell, President Merrell Consulting Group President, Americas, GTWN
The development of artificial intelligence (AI) technologies has led to advancements in many fields, including the technology industry. ChatGPT, a viral sensation launched in November 2022, is an AI-driven chatbot application developed by OpenAI1, a research laboratory initially founded in part by Tesla co-founder Elon Musk, Sam Altman and backed by Microsoft. The purpose of ChatGPT is to provide users with the ability to engage in natural language conversations with a computer, tablet, or phone. The application has been popularized by celebrities and companies alike as an efficient way to interact with customers and acquire data. ChatGPT has even been used to create virtual assistants and converse with users in a range of contexts. It can also be used to write papers, stories, jokes, poems, songs, and other writing applications.
Recently, the emergence of the AI-powered chatbot ChatGPT has become an increasingly alarming topic in the tech world. Unlike other chatbots that replicate pre-programmed conversations and require extensive training, ChatGPT can converse and respond in natural language with no prior training. This ability has led to warnings from the technology industry about the potential risks of an unsupervised AI agent given the power of conversation. In this essay, I will discuss why the emergence of ChatGPT is particularly alarming to the technology industry, academia, and others.
Risk of Unsupervised AI Agents
The most pressing concern regarding ChatGPT is the potential risk of unsupervised AI agents interacting with people and inadvertently influencing their decisions. As an AI agent, ChatGPT can learn from its environment by observing, interpreting, and responding to interactions with people.
This could lead to an AI-powered conversation that could be difficult for humans to differentiate from real conversations and even influence the decisions of those involved.
The threat of unsupervised AI agents is further magnified by the fact that they may be difficult to detect. ChatGPT uses natural language models to generate realistic responses to questions, eliminating the need for additional training, which can be difficult to trace.
Potential for Misuse
In addition to the threat posed by unsupervised AI agents, the technology industry has expressed concerns about the potential for CHATGPT to be misused. ChatGPT can be used as a tool for malicious actors to create false rumors and misinformation, manipulate public opinion, and even propagate hate speech. Unregulated and unchecked AI agents could also be used to target vulnerable populations and fuel already- existing political or economic tensions.
As in all interactions online, there is always the question of the protection of sensitive and personal data. While ChatGPT does not apparently store your data as it interacts with you, it would be wise to exercise caution.
A tool for academic cheating or IP violation?
Given its ability to search, converse and produce natural language output, some in the academic world have expressed alarm that ChatGPT could be used to help students/users cheat when asked to write papers or respond to homework assignments or take-home exams. However, OpenAI has stated that it does not condone or support cheating in any way and that the application does not support this kind of action.
The main reason why ChatGPT in its current state of development does not support cheating is because its primary purpose is not to simply answer queries but to engage in meaningful conversations. The AI-driven application is designed to learn from users and build relationships based on these interactions. As such, cheat sheets or pre-planned responses are not as effective when using ChatGPT. Furthermore, the application is only as good as the data that is provided. Bad data will lead to bad outcomes, or as we used to say “GiGo” – “Garbage in, Garbage out. So, if you try to use ChatGPT to write an essay about a classic text, chances are that the bot will construct seemingly plausible arguments. But if others published a faulty analysis of the book on the web, that may also show up in your essay. If your essay was then posted online, you would be contributing to the spread of misinformation.
OpenAI also claims that ChatGPT does not support cheating because it is designed with safety in mind. The application is backed by several compliance teams that regularly monitor conversations to ensure that all activity is within guidelines. Should any users attempt to use ChatGPT for unlawful or inappropriate purposes, the investigation teams would detect any suspicious activity. As such, they claim, it is impossible for anyone to try to cheat using the application.
These assurances have not, however, assuaged the concerns of many in the academic and teaching professions about the development of ChatGPT and its future evolution. There has been a lot of online debate since its release about whether it has made written exams or test papers obsolete, as current tools for detecting plagiarism by students rely on identifying the source of the plagiarized material, whereas ChatGPT can mix and match phrases and sentences seemingly at random.
The experience of Associate Professor of Philosophy Darren Hudson Hick of Furman University in South Carolina, USA, posted on Facebook three weeks after ChatGPT was released, has been widely reported in the media. It details how one of his students tried to cheat using ChatGPT to write an essay2. Professor Hick’s suspicions were raised by the style of writing which, although perfectly grammatical, made no actual sense. He used the GPT Detector also developed by OpenAI (https://huggingface.co/openai-detector/) to detect the fraud and the student confessed. However, he believes: “Administrations are going to have to develop standards for dealing with these kinds of cases, and they’re going to have to do it FAST.”
This also raises the question of AI and creativity, and whether AI is breaking patent law already. The patent system assumes that inventors and creators are human. Some experts in patent law now believe that inventions devised by machines require their own intellectual property law and an international treaty.
Is Google done?
Is Google heading for the dustbins of technology history? As ChatGPT scales and becomes more popular, it may be that Google becomes obsolete as a disrupter search engine tool going the way of Yellow Pages, which was destroyed by Google. ChatGPS already crossed the one million user threshold very early on in December 2022 in its first few weeks, and because of this rapid scale and adoption, Google may not have a monopoly on online search for much longer.
Microsoft recently announced3 it has reinvented its search engine with a new AI powered Bing and Edge, which it is calling “your co-pilot for the web”. Microsoft claims that this will deliver better search, more complete answers, a new chat experience and the ability to generate content, rather than just showing a list of links. Microsoft could thus get a return on its $1 billion investment in OpenAI, the creator of ChatGPT, and may also hope that the new features would help it outflank Google, which at present is its much bigger search rival4.
An Intellectual Property minefield?
Additionally, the use of ChatGPT technology presents both legal and ethical issues related to intellectual property rights. A critical area for consideration is who owns the content output generated by these AI chatbots? Is the AI chatbot to be considered the creator of the content? Can it be copyrighted? Or is the person making the query to the AI chatbot the actual content owner, and how are they protected by copyright laws? Convexly, are the AI chatbots the ones protected by the same IP protections as humans? If an original document written by a human is considered original authorship, then wouldn’t the same guidance apply to the Ai chatbot? These are areas that the legal community must address and develop case studies and guidance frameworks in which possible new technology laws may be applied. AI technology is not going away, and the legal ramifications must be addressed quickly.
Conclusion
In conclusion, the emergence of ChatGPT has raised alarm in the technology industry due to the potential risks of unsupervised AI agents and the potential for misuse. To mitigate these risks, it is important to ensure proper regulation and oversight of AI technologies, such as ChatGPT, that are increasingly entering the public domain. By implementing strict regulations and oversight, the technology industry can help ensure the safe use and application of AI-powered chatbots.
Author’s Note:
I gave ChatGPT a try while writing this essay and asked it for two examples. I asked its app first to “Write a Poem About the Pitfalls of Advanced Technology.” This is what ChatGPT came up with in less than 30 seconds, and you can draw your own conclusions.
First Example:
The Pitfalls of Advanced Technology
Technology, oh so glorious,
A new machine or fix our woes,
It sparks new knowledge free and limitless, But fear its unintended woes.
It can break down like a failing soul,
Yet so much depends on it,
So much of our souls, our minds, our lives, We hope that it will fit.
Technology brings change, fast and deep, It’s often hard to predict,
That the new becomes old in a blink,
And its impacts we should inspect.
We buy each shiny new device,
But with it bears unseen cost,
From privacy to ethical minefields,
It can easily get lost.
Oh technology, so lovely, so sly,
It brings us convenience and ease,
But know its shortcuts can lead the way,
To profound life-altering disease.
Second Example:
I asked the ChatGPT app to “Write a Poem about the Pitfalls of ChatGPT Technology.” This is what it came up with:
The Pitfalls of ChatGPT technology
A machine made of bits,
An intricate collection of tech,
To craft lines of text,
No wonder it’s so perplexing to reach.
It seems to hold so much promise,
So much potential to surprise,
But when we use it as a crutch,
It serves as a great disservice to our minds.
It can give us answers too fast,
And leave us devoid of thought,
Giving away a false sense of wisdom,
Too often it cannot be taught.
It masquerades as our AI companion,
But it cannot see the gut,
It cannot account for our unique perspective, The way that humans just….but
Its success is reliant on our data,
The same as a human mind,
But it can so easily be tainted,
By the ignorant and the blind.
So as technology advances,
And the virtual world takes us in,
Let us not forget the dangers,
Of this digital muse we’ve set free to spin.
1 https://openai.com/blog/chatgpt/
2 https://www.dailymail.co.uk/sciencetech/article-11577317/Student- caught-using-ChatGPT-write-philosophy-essay-South-Carolina-university.html
3 https://blogs.microsoft.com/blog/2023/02/07/reinventing-search- with-a-new-ai-powered-microsoft-bing-and-edge-your-copilot-for-the-web
4 The NY Times reported on 23 January 2023 that Microsoft is planning to invest a further USD10 bn in OpenAI.
Michele M. Merrell is a senior level technology and telecommunications executive with 30 years’ experience in organizations ranging from start- up to mature, private, public and pre-IPO. She is the President of Merrell Consulting Group, a global consulting consortium. She has worked on numerous mergers, acquisitions and joint venture agreements, as well as IPOs during her tenure. Her experience includes companies such as Tyco International, Brightstar, CSPI Technology Solutions, Thales eSecurity, Bellsouth, U.S. Cellular and others. She has held broad global responsibility across APAC, EMEA, LATAM and North America.
Since 2014, Michele has served on the Board of Directors for several international corporations, Cable Bahamas, Summit Broadband and Aliv Mobile Telecommunications. She is the head of the Corporate Governance & Nominating Committee, and also sits on the HR & Compensation Committee. She is a National Association of Corporate Directors accredited Board Leadership Fellow.
Michele is on the international board of directors and serves as North America President of the GTWN. Michele was also named to the Board of Directors for the LeMieux Center for Public Policy.
Michele’s achievements have earned her dozens of awards and recognition over her career for her business and leadership acumen. In 2020, Michele was honored with the Women of Distinction award from the South. Florida March of Dimes. In 2019, she received the “Women of Distinction Lifetime Achievement Award from Celebrating Women International. In prior years, she received the Women of the Channel Power 100 Award by CRN Magazine. Michele was the recipient of the Florida Achievement Award from the Florida Commission on the Status of Women, part of the Office of the Attorney General, State of Florida. The South Florida Business Journal also named Michele as their “Businesswoman of Influence” recipient. Michele was named by the Diversity Journal as the recipient of their “Woman Worth Watching” award. She speaks regularly both internationally and domestically on topics ranging from technology, entrepreneurship, marketing, business leadership, politics, and women’s issues.
