This post is not off-topic, on the contrary, it's very much on topic, since the topic is what kind and amount of evidence is necessary before we can accept some extraordinary circumstance took place, whether alien visitation or otherwise. If some kind of evidence is to be excluded, it's important to clearly articulate why, even if it's well-trodden ground.I was just about to ask about a similar scenario.
Imagine you heard your neighbors outside talking excitedly one night and you went out to see what was up. You go out to see a giant dull black triangle blotting out the stars for what appears to be at least two miles along each of its edges. You see nine circular, recessed cavities underneath it that contain a warm globular amber glow. Whatever it is is slowly passing directly overhead and, of course, it’s silent. After watching it for 45 seconds or so and attempting to film it with your phone it suddenly shoots straight up at impossible speed and disappears. You say to your neighbors, “did you see that?” and they say of course they did. It’s the most breathtaking thing you and your neighbors have ever witnessed.
You reluctantly post your less-than-ideal video on metabunk and we predictably say: inconclusive. Could be drones plus optical illusions. Could be doctored. Regardless it’s just a video. You explain your neighbors’ reactions, link to their even shakier nighttime videos. Metabunk confirms what you already know: inconclusive. Proof of nothing.
If the intellectually honest thing to do is admit it almost certainly wasn’t anything otherworldly (eyewitness testimony and videos are unreliable) could you, as baffled witness, genuinely accept that conclusion? When describing what you saw to others, how would you frame it? Or, since you can’t have known what you saw, would it be wise to avoid mentioning it altogether?
Note: Edward Current, if this is too off-topic, say the word and I’ll gladly delete this.
Can you, as a witness, be convinced that "you know what you saw"? Certainly, though in the vast majority (all?) of such cases witness do not in fact know what they saw -- they were convinced by a mistake of their senses, an illusion, a phantasm. They think they know what they saw, but what they actually saw was Venus behind the trees, or airliner landing lights in the distance, etc. Notice also that in many cases where people "know what they saw" and there are multiple witnesses, the witnesses disagree about key details concerning the sighting (the tic tac case is a notable example).
But take a situation like you described where it's truly unambiguous what's going on, there are multiple witnesses, who were not on drugs etc. and they all agree about what they saw and it matches what you saw. Should you, the witness, be convinced? Yes! The problem with witness testimony is not that it's unreliable, though that's a big part of it (which I'll get to in a moment). We are, after all, eyewitnesses to the recorded evidence, and we can be convinced on the basis of that evidence. Science has a largely positivist ethos which means every piece of evidence will ultimately be filtered through human senses, so the reluctance of skeptics to accept eyewitness testimony has nothing to do with some sort of "anti-human" bias as some have claimed.
The problem with eyewitnesses is their experiences are not repeatable. You know what you saw, but nobody else does. We can't verify it, we can't analyze it, we can't do anything with it. Nobody has access to your subjective experience and there are plenty of reasons why you might claim to have had such an experience, but have been mistaken, or exaggerating, or confabulating, or even outright lying. From a perspective outside your own head, all of those alternatives would look essentially identical in terms of observable effects it produces (namely the sounds coming out of your mouth or the keys you press on a keyboard), which means that each of these a priori vastly more likely alternatives is still more likely to be the correct one. Should, then, anybody else be convinced? The answer is no.
There is no paradox here: information is often subjective and asymmetric, which means that the kinds of mechanisms that can explain observations and their relative likelihoods are themselves subjective. This is something Bayesians understand well:
(this cartoon of Bayes' theorem comes from this presentation, which is a very good introduction to the subject in the context of skepticism. for a deeper dive, see this book draft by ET Jaynes, the godfather of the modern perspective on the subject in the context of science and rigorous reasoning).
This is a fundamentally subjective calculus, that is done differently for each person depending on what evidence they have and what information they know. You, as a witness, can discard any number of competing explanations in that denominator there -- you know you aren't lying or confusing an obvious giant spaceship for venus, but for everybody else those competing explanations are well in play and they drive down the probability of actual flying triangle down to nothing. You believe you saw a flying triangle, we don't, and we're both evaluating evidence rationally. Perfectly possible.
This also explains what is meant by the unreliability of witness testimony. For claims of ordinary phenomena, the prior probability (how typical our explanation is in the above cartoon) is high (by definition) which means the probability the phenomenon happened is high. So if you tell me you saw an airplane come in for a landing when driving by the airport, I will likely accept that claim as true. If you say you saw a flying saucer, I will evaluate the claim differently. And, of course, while in most of this conversation I've been happy to assume that you do in fact know what you saw, in practice I also have to assume that you may simply not have thought of an alternate explanation, and I have to consider those as well. If you are evaluating evidence rationally, you should incorporate those alternate explanations in your own probability calculus.
I hope this post is of some use in contextualizing why "direct human interaction with our world" is adequate most of the time but fails to meet the bar for convincing skeptics of the reality of any extraordinary phenomenon.