Are AI Chat Bots Inherently Harmful? One Study Says “No”

The age verification debate has morphed into roping in AI chat bots. We found one study on the impacts of AI.

One of the most terrifying things about age verification is the fact that what gets roped in keeps expanding. It is no longer tied specifically to dedicated porn websites with a certain percentage of pornographic content. It has long since expanded to include social media, search engines, VPNs, app stores, game platforms, and now, more recently, for cell phones in general. While most of this is just pushes for inclusion, Canada has also expanded it to include AI chat bots in their latest push.

If you have followed the debates for any reasonable length of time, the talking points are all too familiar. When the government came for your social media, the talking points were that social media is inherently harmful and “addictive”. Therefore, children should be barred from social media altogether to protect their “developing minds” and to prevent social media from “rewiring their brains”. The talking points are complete and utter bullshit and science has long debunked these myths. We’ve seen this in study after study after study after study after study after study after study. Yet, for the fraudsters and con artists pushing these narratives, they’ll continue to lie through their teeth and argue that the science is in and that they were right all along. Even worse is the financially motivated biased mainstream media giving these con artists the biggest bully pulpit they can offer to spread their lies and disinformation. In fact, the mainstream media will even go so far as to defend this practice by telling people who call them out that it is not their job to fact check. They know they are publishing lies and they will not only refuse to apologize for it, but pretty much argue that they are proud of the fact that they are deceiving people.

In fact, the mainstream media is also doubling down and arguing that giving someone access to social media is the equivalent of letting a child use a table saw and even arguing that social media is killing children – as if to say that accessing the home page of Facebook could mean a gun will phase through a persons monitor and shoot you in the head or something crazy like that. The absurdity of their talking points is that far off the charts at this point in time. How anyone even believes this crap is astonishing to me.

Yet, mixed into these debates are AI chat bots. This move is largely coming out of left field, but the talking points are the same. The mainstream media and politicians are arguing that AI chat bots are inherently harmful, and, as such, we must ban those too. It’s a fair thing to say that AI chat bots aren’t as well studied as social media. This makes busting the myths pushed by mainstream media harder. Just the same, it also means that the science is not even close to being in to determine what action needs to be taken and, as such, those pushing for this are on shaky grounds as well.

If, this is the government and the media taking advantage of a tragedy to push a political agenda. Much like how the media and the government tried to take advantage of the Yellowknife wildfires and evacuations to push the Canadian link tax, the government and mainstream media are taking advantage of the Tumbler Ridge mass shooting to push this particular agenda.

For those that don’t know, the Tumbler Ridge mass shooting was the worst mass shooting in Canadian history. I touched on this here, but one of the things the media did early on was go to Tumbler Ridge and ask residents about their experiences. It’s probably the only time they did Canada a solid because the stories that came out really painted a picture of how badly things have slid in terms of community supports from the governments perspective. One resident told the media that Northern Health (the regulatory body that covers British Columbia) only has ambulance services for certain hours of the day. In other words, if you plan on having an emergency, you may want to wait until it happens during the hours of operation for ambulance services.

Indeed, before the mass shooting, this was at least talked about with reports of frustrated residents demanding better services from Northern Health. Since the mass shooting, however, there has been nothing but radio silence on this issue from the medias perspective.

Another issue that came out of this was questions on whether or not mental health services were sufficient in the community. Was appropriate funding in place? Is there sufficient awareness of such programs? Are they easily accessible? On this front, the biggest story was that the government brought in teams of mental health experts to help those dealing with the tragedy. Otherwise, the recent reports I was able to find suggest that mental health services for Tumbler Ridge continue to have a massive funding gap. In fact, the mother of the shooter actually attempted to access mental health services and was ultimately unsuccessful:

In a statement shared to the Globe and Mail, the shooter’s grandparents said the family had tried unsuccessfully to get help for a variety of mental health concerns.

The mother of a girl still in hospital told reporters she knew the shooter’s mother and her struggle to get support and adequate care in the small community of just over 2,000.

“It’s about mental health. It’s about a lack of resources,” she told reporters.

Like, how is this alone not a problem that needs addressing? It’s plain as day how obvious this needs to be addressed.

If anything, the mass shooting should have been the kick in the pants to try and provide better services in that community. Both physical health and mental health services are woefully inadequate and residents have been demanding change for quite some time now, yet their calls have been going unanswered for a considerable amount of time.

Yet, the main thing that the mainstream media is fixating on is the role OpenAI and ChatGPT had in all of this and whether more regulations are necessary. Why go through all the hassle of trying to fix a bunch of systemic problems when you can just blame technology and dust your hands of this whole thing? It’s fair to say that banning chat bots isn’t going to address the shortfalls highlighted above, yet this is the direction the government has taken.

It’s all very frustrating, but there are those who will be wanting to know what actual impacts chat bots have on younger people. Recently, TechDirt published and article railing against John Oliver for completely dropping the ball on the debate about AI chat bots. The criticisms were very valid, but one thing that did stand out to me was a reference to research conducted on the use of AI chat bots. That study was from JAMA Network and the results were actually pretty interesting:

Of 2125 individuals contacted, 1058 responded (response rate, 49.8%; 37.0% aged 18 to 21 years; 50.3% female; 13.0% Black, 25.2% Hispanic, and 51.3% White) (Table). In total, 13.1% of respondents reported using generative AI for mental health advice (Figure), with higher rates among those aged 18 to 21 years (22.2%). Among users, 65.5% sought advice monthly or more often and 92.7% found the advice somewhat or very helpful. All results percentages are weighted.

Now, I’m not saying that a chat bot is a good place to get mental health advice. I would advise speaking to a human professional for stuff like that. Still, these findings show that the narrative that chat bots are inherently harmful aren’t exactly holding up to scrutiny.

Now, I’m under no illusion that actual science is really going to make a difference in this debate. The push for age verification has long been more of an appeal to emotion rather than common sense and reasoning. Anyone who dares to question the narrative are cast aside as either stupid or not worth listening to. We’ve seen this in the Online Streaming Act debate and we’ve seen this in the Online News Act debate. We’re seeing this in the warrantless wiretapping debate as well. Age verification is shaping up to be no different. Still, at least we are able to compile evidence that counters the narrative and when things invariably go sideways, we can smugly say “we told you so”. Wouldn’t even be the first time that’s happened.

Drew Wilson on Mastodon, Bluesky and Facebook.