Raw LLM Responses

That thing is behind a paywall, so I have no idea what the article says. Does he propose a mechanism here? Because as I see it, the difficult part of "a biological attack" is not theory. Just because an AI, or someone else for that matter, tells you how to do it, doesn't mean you can actually just sit down at your kitchen sink and do it. The "know how" is not the difficult part. For example, in theory it's not particularly difficult to engineer a very deadly flu strain: You just need to cultivate a lot of succeptible human tissue in cell culture, infect it with a flu virus, select the most virulent strains, repeat that process, and you are well on your way (at least in theory, if everything goes well, and nothing goes wrong, hahaha as if). Dear "evil person" planning a bad biological attack, I expect a thank you note, now that I have told you all about how to do that. The problem is in the execution: For all of those steps you need some special equipment. At no point in the process will DIY cut it. If you want to infect a human cell culture with a certain virus, and be sure that it's infected with the exact virus you want, you have to do that "isolating a virus" thing first. You need a whole lab for that. And it has to be a rather well equipped one. It's also not very easy to keep a human cell culture going in the first place. Not enormously difficult, but definitely not a DIY thing. You need some specialized equipment. Oh, and if you want a human cell culture, you first have to get yourself a human cell line from somewhere. You can't make that yourself either. It's also not very easy to work with viruses which could infect the actual humans working with them. You need, once again, some specialized equipment to prevent that. And all of this is the problem: Your next door terrorist is not well funded enough to build themselves a lab specialized in producing bioweapons. If he is well funded enough, chances are good that a lot of people will take note that someo