I hope you fact checked the Wikipage, the Canadian Youtuber "Not What You Think" and the German Youtuber "Sea Power Reports" that your AI cited as sources. Of course, given your expertise on Naval engineering, explain why the President is against EMALS, which you consider irretrievably unreliable, but he is strongly supportive of so-called Rail Guns on his Golden Battle Ships The problem with your reliance on AI is that it cherry picks statements to support your Prompt, which you have decided not to share with the group. Here's another AI generated statement: Prompt: Provide a concise argument, limited to one paragraph, addressing the perceived reliability issues of EMALS in comparison to steam catapults on aircraft carriers and why these are overstated Now, same AI, different prompt: Prompt: Provide a concise argument, limited to one paragraph, addressing the perceived reliability issues of EMALS in comparison to steam catapults on aircraft carriers and why EMALS is unlikely to achieve operational reliability. Highlighted are contradictory statements. One says that EMALS is inherently simpler and easier to maintain. The other says that EMALS is inherently more complex and difficult to maintain. The AI is goal focused; its goal is to support my leading question. The problem with AI is that it is stupid. If you have a background in defence engineering and design, you can construct a prompt that can give a truly informative response. As the SME, you act as a conductor, reviewing, rejecting, assembling the response, often step by step. If you are using AI to tackle areas you know nothing about, then it runs into trouble, unless you possess inherent skills as a researcher. In your prompt, the two substantive sources are a pair of anonymous Youtubers. If you provided a response saying this is what these Youtubers/Influencers are saying, that would be a reponse with low credibility. But when its an AI response, that's suddenly more credible. AI changes the research process. You straight away start with lots of sources, which you now need to study each of, appraise, and it might take you down a rabbit hole. Conventional research though starts with no sources. Somewhere along the line you have to form your own hypothesis, and test that with available data. Ultimately, when it comes to head to head comparisons of defence systems, the people that actually know are not talking. And that's because that is operationally sensitive information, and they likely would be prosecuted for divulging essentially state secrets, whether that's EMALS is terrible, but don't tell the other side, or that EMALS is fantastic, but don't tell the other side that.