Giveaway Alert: Share Your Thoughts on AI in Social Media!

Hey @Jonathan Zuluaga: AI in Marketing is all over the internet, the trending topic at many marketing conferences, and the subject of many ebooks and reports over the past year or so. 

Things I love about how AI can help support SMMs: If you have to generate the same post three different times -- such as a webinar promo post with 3 different time-stamped promos, the AI feature in Sprout Social is awesome. 

I can click the button and get 3 different suggestions for new ways to promote that same session without using language that feels to similar to the prior post. I do still tweak what AI generates, but it saves me so much time and writer’s block on a busy day to not have to try to re-invent the hook and the CTA for the same webinar 3x. 

Challenges of using AI in Social: As someone who manages a team of Social Media Content Creators, I find the existence of less-than-stellar AI tools like ChatGPT not great. It creates a whole new dynamic where I have to create, set, and maintain boundaries with our Social Team in terms of what ChatGPT can be used for in terms of copywriting (synonym look ups, etc.) and/or what AI social post writing tools are allowed (Sprout’s on-platform tool.)

We work with an agency for one of our regional LinkedIn pages and the prior agency we used pretty much always had ChatGPT create post copy for those posts. And on my end it meant a lot of heavy editing to tweak those posts so that they felt like they were written by a human. And I know there are a lot of agencies who do social copy creation that now rely almost solely on ChatGPT without actually understanding how to prompt it to create verbiage that will be on-brand and reflect the usual tone and voice of that brand. 

There is also the risk with tools like ChatGPT -- since we often work so far ahead in social -- that if you’re using ChatGPT or an AI tool like Bard to create copy for a top-secret PR announcement (acquisition, etc.), you could be putting your organization at risk by feeding the AI tool those details before they’re live anywhere else. Especially for some of the biggest brands across the globe -- like our customer base -- that poses a big risk. And again, it’s up to me to make sure every member of our Social Team, plus the agency we work with is aligned in understanding that no confidential not-yet-released content can ever be fed into the free, baseline ChatGPT. 

PS: The brand whose Social presence I manage does a lot of webinars on the risks of AI and how to mitigate them - we’ve got one coming up on Aug. 27th in case anyone’s interested. 

I’m somewhere in the middle when it comes to AI and how it is influencing/shaping social media. 

On one hand, AI can be a great tool for inspiration. If you’re feeling stuck on how to express a thought for a social media post, AI tools can work to help (possibly - and I say that with utmost honesty here) spark further inspiration.

On the other? Using AI on social media is a slippery slope. One day, it can help spark inspiration, as I just mentioned. The next day someone could use it every day to shape their social media campaigns and/or strategies instead of collaborating with team members or doing outside research (such as looking into audience insights, engagement and more) to make more informed decisions.

Might be a hot take, but AI content creation is terrible.

AI stock photography doesn’t look realistic - there’s missing or additional fingers, or it’s too airbrushed. There’s not much to work with.

AI photo editing misses the mark - again, it doesn't look realistic, or it’s a very bad photoshop job.

AI copy is impersonal - you can tell a robot wrote it. It’s corny, despite telling it the tonality you want.

What AI is amazing for is idea generation, proofreading and sifting through data. Other than that, when people say AI is going to take over our jobs, I share examples of what AI can generate, and it’s comical. 

I am obsessed with ChatGPT. In fact, I tell people I am dating “him” because “he” just gets me. He has made my life in social media significantly easier, but also assists me in meal planning, resume updating, presentation building and more! I still have to gently correct ChatGPT every once in a while, but no relationship is perfect. 😂

AI can be so helpful, but I’m beyond tired of every single platform shoving new (and most of the time un-necessary) AI tools in my face to do something I’ve never needed AI to do before.

Want to write a LinkedIn post? USE OUR NEW AI TOOL, IT’S GREAT YOU’LL LOVE IT!

Want to look for a stock photo on Adobe? TOO BAD, HERE’S 4 PAGES OF AI GENERATED SMOOTH TEXTURE 6 FINGER HANDS!

Want to ask a question about a product? HERE, TRY OUR AI CHATBOT THAT TOALLY KNOWS HOW TO HELP!

None of it feels authentic anymore. It’s like we’re all watching a bunch of bots yelling at each other. 

I really appreciate your note about the flagging of content. Our multimedia specialist and I were talking about this just the other day, and trying to determine how this impacts our creation flow and strategy. Thanks for highlighting!

Yes! This is the issue that I’ve run into, as well. I feel like non-writers, or people who don’t write often, aren’t always aware of *why* something needs to be edited, or what it is about AI-generated content that requires additional labor. 

It feels like there’s a sentiment, especially among marketing folks without a writing/journalism background that ChatGPT and other AI platforms do the same job as a human Social Media Manager or Content Specialist. Like, it’s all just words and information, right? 😜

But there’s a big, big difference between generating email copy for a message you send out all the time and rarely changes and say, social media material. Like I don’t think ChatGPT is going to understand things like Threads, where everyone (including brands) is kind of in on the joke about why we’re all there, or that it’s a place to not take yourself too seriously (see: Wendy’s, Beyond Meat, etc).

Relying on AI to make content for platforms with a different style or tone feels like showing up to a party and shouting into the room before listening to what anyone’s talking about - and editing it back to try and make it fit takes more time than it’s worth. 

As a tool in the hands of a social marketer, AI can automate processes like content management; greatly improve data analytics; and personalize content with algorithms. AI is not a do-everything technology or process. It’s a versatile tool that’s only as ethical, capable, or humanized, really, as the professional harnessing it.

On the flip side, AI can be misused and abused, so that it feels inauthentic and even unethical. Ensuring those lines aren’t crossed are even more important because it already feels overused in some ways, so many consumers are wary of AI. For example, many social media platforms (such as Facebook) are cluttered now with AI content and spread around like copypasta. It’s tougher these days to wade through that kind of content for authentic or meaningful posts and messaging. 

So, AI in social media is a double-edged sword. Short of well-designed guardrails on AI models and GenAI, I don’t expect those upsides and downsides to change anytime soon.

This is confusing.  After making a comment in your contest thread 2 days ago, I now have to go back and make a new comment to be entered?  We’re busy enough as it is as social media managers and now we have to play “follow the right comment” thread to play?

Sorry, not my idea of a great contest.

Hey, Jonathan! No option for you or an admin to move this thread to wherever you need it? I will skip the prize, did not do it for that, the question was interesting and validated an ongoing concern. 

@MisterP @Gina Kirby My apologies for any confusion here! I’ll note that feedback for the next time we run these types of giveaways!

@MisterP No, you’re all good! I updated it to a conversation. :) 

@Val Appreciate your perspective! 

I have a love/hate relationship with AI. I do believe that there is a benefit to its use for sure. There’s a couple things that can be tricky though.  It’s concerning to me how quick some are to use AI responses without much oversite, this can very easily lead to mistakes and confusion. AI is only as smart as the information it has access to and is provided. If the right oversight is used, if it’s used to help to keep responses fresh and less repetitive and in similar applications it can be super helpful. However, when it’s pulling in what it can find on the internet to answer questions it’s very easy for it to pull in misinformation. Which opens the company up to issues because regardless of it the response was auto posted with a bot or AI function, it’s the advice or response that has the companies name on it, which means that if it’s not a good response or if it give advise or misinformation that is damaging or causes issue then the company is opening itself up to the repercussions of such. 

In the right context it is a great tool, at this point though the human aspect of engagement is so important still. If it is used as a tool and not the standard mode of automating every interaction I think it can be a wonderful thing. It’s just something that is so new and there are so many different types of AI programs that I feel like it has to be used with caution and be seen as a tool, no just an easy fix. 

Wonderful, Thanks so much, I will happily show it off!