(Aside: I'm a published neuroscientist and computer scientist with a very active but currently non-professional interest in how the two fields are colliding to cause huge problems. Arguably, my two main fields are in somewhat direct conflict as we move into a post-social media IoT/AI world. The only way I can reconcile them and my expertise in both is that computational neuroscience is going to be a huge part of the future of diagnostics/epidemiology/aetiology/treatment in psychiatry -- and that's my goal in life.)
With all that said, you're not exactly wrong, particularly from a "mainstream" perspective.
We're starting to at least ask the right questions in the literature. And we have, for example, long since established that social media is basically terrible for everyone, especially kids, in all but the smallest of "doses".
One of the problems in that area is that the mainstream media is not exactly desperate to report on problems associated with social media. Compared to other far more sensationalist "health scares" of the kind stereotypically perpetuated by the Daily Mail, reporting for laypeople actually vastly understates the quantity and quality of the data we have on social media use and its correlates. It's like they don't really want to report on the matter (you can speculate as to the reasons; I have my own theories).
As far as I'm aware, we're beginning to assemble a solid base of convincing evidence that giving your young kids smartphones and tablets - quite aside from obnoxiously spoiling them - is not good for their development. Many effects are obviously highly dependent on how they use the devices - limited exposure to some genres of video games has positive effects on abstract thinking, executive function and reaction time with predictable grey matter volume alterations on neuroimaging - but most contexts are negative and there are concerns around things like opthalmologic development under heavy exposure to artificial light (with obvious secondary effects on circadian rhythms). It's not unlikely that almost all of the generation being born right now will need glasses before 30.
My advice on that for anyone who has kids would be to carefully control what they do (games are actually much better than social media) and to limit their overall exposure if you are going to give them such devices (and I know there's mass peer-pressure going on where seemingly all the other parents are being irresponsibly overpermissive). I've been saying for years now - only half joking - that when I have kids, they're not getting on the internet until they write their own TCP/IP stack from scratch.
Academics are just fallible humans at the end of the day and most people I meet in CS and in neuroscience simply don't want to think about the dystopian future we're currently driving towards or the seemingly-distant implications of their own work. And it's easy enough to ignore for the vast majority of people, even in a field like AI. For example, if you're working on an efficient binary image classification algorithm, you're probably not thinking about how part of your idea could one day be used to select targets for drones to auto-kill on sight. Only a handful of people ever get a macroscopic view of putting all these tiny pieces of research together for anything outright nefarious.
Anyway, I'm going on a bit so I'll stop but I hadn't even reached the clusterfuck of controversy that is something like 5G: a whole lot of very powerful people have a vested interest in making sure the roll-out succeeds and there's sort of an unspoken pressure on researchers not to push too hard, which is why we're getting 5G all over the UK when there hasn't yet been a single high-quality long-term trial of human exposure. There are a lot of people trying to quell their nerves and/or conscience right now by repeating the (incorrect) mantra that "non-ionising radiation has no effect on humans".
