Hello Moeen,

In my opinion, Grammerly and Quillbot have been used to assist, improve, and enhance the language only. ChatGTP is different because it thinks, writes, does anything instead of you. I think it can be used as a tool to help and improve our information, but we should be aware to avoid: the plagiarism and violating the ethical rules and guidelines.

In addition, it must take into consedaration that AI, including ChatGTP, works according to the try and error method and comparison with what the users and programmers entered before, and modifies the results continousouly. So, AI is not always accurate 100%.

Creating ideas? At the current level of maturity, maybe, but probably not, unless the tool is regurgitating an idea input by another user. This is an interesting question, which if true, has profound implications. This is where ethical behavior assumes even greater importance. This is an intimidating capability if and when it manifests. The military definition of a threat is capability combined with intent. I'm OK with AI having capability. If it becomes sufficiently aware to the point where it manifests intent, we need to have a very, VERY well developed sense of peril.

In my opinion, the use of AI as a tool to develop ideas has become indispensable from a productivity perspective. As long as you're smart enough on a particular topic to "drive the tool", your productivity can skyrocket. A good analogy is that AI-based tools are akin to a precocious teenager with zero life experience. AI tools exhibit massive amounts of raw horsepower that need to be properly guided, managed, and VERIFIED, as a lot of the output is just simply incorrect. Blatantly incorrect sometimes. Every bit of output has to be rigorously examined and assessed for accuracy. My point is if YOU drive the tool for purposes of productivity it's fine and ethical. This of course, assumes the individual driving the tool has ethics. If you're sufficiently unaware, and fall into the trap of letting the tool drive YOU, all sorts of horrible results and potentially damaging, unethical results may ensue. This is a sure path to chaos and one we'll undoubtedly be seeing a lot more of. IMO, the average person won't be able to tell the difference. (Reference Yuval Harari). Danger here...

A a personal level, and in many other circumstances the concept of tradition has a lot of value. When applied inappropriately, usually by those who fear progress, tradition can degenerate into "being controlled by dead people". Tradition helps us feel safe because that's the way things have always been done. AI tools throw the equivalent of a psychic hand grenade into not only our minds, but directly into the concept of tradition. It's been said that "the concept of progress is a conceptual shield that protects us from the fear of change". There's a lot to unpack here regarding the direction and impact that AI will have, but I don't want to digress so I'll move on.

Oger's response mentioned plagiarism. In an age where a kaleidoscope of AI-generated content is just beginning to bombard us, I wonder if as a species we can keep up. I don't know. In fact, the more I think I know, the more I realize I know nothing. I can have an opinion though, and I can learn. I can modify my opinions based on new FACTS that make an old opinion or position untenable. I use GPT-4 and Google Bard on every idea I have. Every single time I use these tools it's like I had a conversation with a really smart friend who sat in my office where we brainstormed an idea and figured out all the ways something would NOT work, and sometimes figure out ways in which it will.

Wisdom is in short supply. I personally have a sense of excitement about a future that includes AI/ML. I also have a very well developed sense of peril, because I understand the shortcomings of some of the people who will be leveraging these tools. The future isn't coming - it's here.

Thoughts? Comments?

Moeen Al-Makhlafi ,

Legality and ethics are completely separate issues

Legality

In the US this is being tested out in a lawsuit against Github's CoPilot [1]. Note that the legality depends on the jurisdiction and what may be determined in one country may not apply in another.

It is further worth noticing that decisions in court may not be a one size fits all within a jurisdiction. For example, the coverage of copyright may not cover databases where generic information is available (that is, there was no creative input even though it took substantial effort to gather).

Ethics

Whether to benefit from other peoples effort and creativity may be debatable. One point is that it is totally unfair since a system is merely permuting the inputs to generate an output based on the data given. This presents a damage to the livelihood of creative people and those whose work is being automated. The opposing argument that I have seen is that the machine is basically doing the same thing as a person(this is a sweeping argument that may be too broad and ambiguous on several fronts). The argument continues with stating the fact that there is progress and that people will need to engage in retraining to get other jobs.

References

[1]https://www.saverilawfirm.com/our-cases/github-copilot-intellectual-property-litigation

The legal and ethical implications of using an artificial intelligence tool like "ChatGPT" to gather information and generate ideas depend on how it is used and the specific context. From a legal perspective, as long as the usage adheres to intellectual property rights and data protection regulations, it can be considered legal. However, the ethical aspect requires careful consideration. Using AI tools to gather information can be beneficial if the sources are reliable and properly attributed. However, blindly relying on AI-generated ideas without critical thinking and independent verification may raise concerns about plagiarism, lack of originality, and academic integrity. It is important to strike a balance between leveraging AI tools as aids and maintaining responsible research practices that uphold intellectual honesty and contribute to the advancement of knowledge.

As long as you double check the information and mention the reference as (this information is a result of a conversation with chatGPT (for example)), then I think it is ok to use these helping tools to accelerate the scientific research process unless otherwise stated in your university or country regulations and laws.

As a further note:

Information gathering is not part of ChatGPT or any AI tool, it is part of a crawler. The job of the crawler/spider is to gather the data, then it is processed ( take out the HTML) and then put in a format for ingestion into the first stage of the architecture(e.g. Word2vec).

The legality of crawling is being called into question for some time (LinkedIn has some interesting cases), but recent debate on ChatGPT has brought the subject again but under two different aspects:

1) data harvesting through crawling

2) data processing of the harvested data and disclosure

These are completely separate issues that must be treated carefully