"just need to be careful to not use AI where more traditional methods would be a better choice"

Yeah, this danger always exists. As a comparison - when CGI became a thing where was a race to replace all practical effects with CGI. But while you can CGI things you cannot create with practical effects, practical effects still have some looks and certain reality to it that is hard to replicate with CGI. New technology should expand and augment existing tech, not just simply replacing it.

I'm of the reverse opinion, comparing to 200 years ago is setting the bar low and leads to no sense of urgency to fix very real problems we have today. But perhaps it needs to get real bad first before people start taking action. People not being able to imagine how horrible it can get is what leads to inaction.

Over 320 million in acute hunger, oh well in the 1800s many more people starved to death. 
Economic sanctions killing 560K yearly (half of those under the age of 5), oh well the plague killed far more.

Unbridled growth from the industrial revolution (also responsible for 15-22 million deaths WW1, 70-85 million WW2) has led to where we are now. Sure many good things came from the industrial revolution, but also capitalism which is now ruining the world. Capitalism is much older yet the industrial revolution cemented industrial capitalism which we are stuck with now. 

Team ambivalent

I see a lot of areas where it is useful, but also how it can be abused by individuals and corporations alike. Off the top of my head:

Benefits

* Productivity Tools: like sorting through vast amounts of information with parameters to guide.
* Search engine competition: AI often proves superior to google for finding the sources you need. Although, it depends on the model, as some models screw up links.
* Research: AI can vastly increasing the speed of research by allowing you to sort and categorize hundreds of studies in a matter of an hour or two, where before it would take a day to get through around 40 to 75 (looking at abstracts, methodologies, and key terms).

As a note, though, I'd still take spreadsheets over AI as the most powerful productivity tool. And this is because spreadsheets are precise and only mess up when human error is involved, while AI reliability is so-so. The other thing is that errors from AI are much easier to miss than spreadsheet errors when processing information.

Risks

* Stunting Intellectual Growth: People will use it as a crutch rather than practice essential skills, like reading comprehension, ultimately lowering intelligence.
* Abuse, Addiction, Social Maladaptation: we already see this being observed, and have seen a similar impact from social media abuse & addiction, especially on Gen Z and younger millennials.
* Erosion of Artist Rights: Further erosion of the rights of artists and media creators by enabling people of lower moral character to violate their copyrights and trademark rights. While some (especially wealthier creators) are alright giving their art for free, it is not a good thing if every artist is no longer entitled to the profits of their work even if they didn't consent to sell it. But this isn't limited to regular cynical people who flout the social contract, corporations have been engaging heavily in this (including video game companies) which means that many creators are out of a job while AI keeps replicating their style again and again. We have seen a massive decline in job opportunities for creators since AI became a factor.
* Bots: This has already proven true for years, and have mostly ended comment sections and open and anonymous social media platforms. AI-powered bots are becoming more sophisticated and more easily mimic human behaviour... it used to be very obvious to spot AI bots because they all wrote with similar syntax and language, but now there is significantly more variety.

I think AI risks are only relevant when considering certain sorts of people and organizations: young people are at greater risk, but also older people with intellectual deficiencies, and selfish/entitled exploiters who use AI to exploit the creations of other artists. Basically, creators are losing their autonomy at an even more rapid pace, exploiters are taking the creations of others without compensating them to instantly generate Frankenstein copies; and there isn't much of a lobby these days for the rights of creators; AI companies basically squashed them. And because AI is becoming increasingly sophisticated, it is becoming easier for customers who want human (not AI) creations to be duped by shysters masquerading their product as authentic human created work (basically, it's inflated the creative counterfeit market).

Closing Thoughts

The long term effects might lead to an increase to the overall conformity and stagnation of art → our culture will become less creative. Are we already seeing this creative decline? What sort of operational definitions can we use to define creativity, for example (and these operational definitions are just off the top of my head, no real deep thought):

• Authenticity spectrum: like, do they engage more in self expression or are they engaging in meme-speak?
• Originality spectrum: are they doing something original to stand out from the pack, or templating an existing trend to catch the bandwagon?
• Thematic exploration spectrum: are they exploring themes in their work, or are they copying a structure without any mindfulness to any original takes on the themes?