AI Has a PR Problem

**Nathaniel Whittemore** (0:00)
Today on The AI Daily Brief, Why AI Has a PR Problem. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.
All right, friends, quick announcements before we dive in. First of all, thank you to today's sponsors, KPMG, Blitzi, Rovo and Robots and Pencils. To get an ad-free version of the show, go to patreon.com/aidailybrief, or you can subscribe on Apple Podcasts. And if you are interested in sponsoring the show, send us a note at sponsors at aidailybrief.ai. Welcome back to the AI Daily Brief. Today, we are talking about a subject which I'm sure will be not a surprise at all to any of you. We are talking about AI's PR problem. There are a number of things going on and stories from the last couple of weeks which all point in a very similar direction. So today, we are going to talk about those stories, what I think is at the root of these challenges, and some very first nascent thoughts on what we can do about it. Let's talk first about this Edelman study. Coursera founder Andrew Ng wrote, Separate reports by the publicity firm Edelman and Pew Research show that Americans, and more broadly large parts of Europe and the Western world, do not trust AI and are not excited about it. Despite the AI community's optimism about the tremendous benefits AI will bring, we should take this seriously and not dismiss it. The public's concerns about AI can be a significant drag on progress, and we can do a lot to address them. According to Edelman's survey, in the US, 49% of people reject the growing use of AI and 17% embrace it. In China, 10% rejected and 54% embrace it. Pew's data also shows many other nations much more enthusiastic than the US about AI adoption. Positive sentiment towards AI is a huge national advantage. On the other hand, widespread distrust of AI means individuals will be slow to adopt it, valuable projects that need societal support will be stymied and populist anger against AI raises the risk that laws will be passed that hamper AI development. So let's talk about this trust study. Edelman publishes their trust barometer every year and this year, surprise, surprise, the big theme was AI and kind of frankly AI consternation of the type that Andrew was just talking about. This survey was conducted very recently, October 17th to the 27th. And boy is this all telling a single story. Edelman's headlines include one, that globally rejection for AI outweighs enthusiasm with US respondents more than twice as likely to say they reject the growing use of AI than they are to embrace it. Even beyond AI, enthusiasm for innovation is not guaranteed. Trust in AI generally lags behind trust in technology. Now, at the risk of being overly reductive, there is a very clear East-West divide here. Although frankly, probably a better way to put it would be developed economies versus developing economies. The survey interviewed people in five countries, Brazil, China, Germany, UK and the US with at least a thousand interview respondents per country. In Germany, the UK and the US., a significantly higher number of people said they rejected AI versus embracing AI. In Germany, it was 42 to 16 In the UK, it was 46 to 18 And in the US., we had the biggest gap, a 32 point difference with 49% of people saying they rejected AI and only 17% saying they embraced it. In Brazil and China, it was the opposite. Brazil had 24% saying they rejected AI versus 35% embracing it, and China had 10% rejecting it and a full 54% embracing it. Edelman found a big income divide, with people who were lower and middle income being more likely to say that AI would leave people like them behind than those in the top 25%. Although in the US., the numbers were high across the board, with even high income folks seeing 47% say that they fear that AI would leave people like them behind. This fear of getting left behind is, I think, one of the key issues that we're going to have to contend with. Unsurprisingly, young people have more trust in AI, but US young people are still distrustful, with only four in 10 US young people trusting AI. And folks in Germany, the UK and the US are very skeptical that AI is going to help any sort of issues from climate change to work life to mental health to political polarization to poverty. In one bit of good news, there is a correlation between people being more informed about AI and having higher enthusiasm, meaning in other words that the more we have people engage with it, the more we might have a more productive conversation. This reminded me of a report that I saw earlier, as tweeted by Business Insider reporter Brian Metzger. Senator Josh Hawley, one of the biggest AI critics in the Senate, told me this morning that he recently decided to try out ChatGPT. He said he asked a very nerdy historical question about the Puritans in the 1630s. I will say, Hawley said, it returned a lot of good information. Another bit of, I guess, good news, although sort of bad news, but maybe good news is that a lot of the problems with AI are perception problems rather than things that people have actually experienced. For example, among those who reject the growing use of AI, still only 18% said that they personally had had bad experiences with generative AI versus 70% who said they had not. In general, when it came to why people weren't willing to use the tools, motivation and access and intimidation, while prevalent, were less common than general trust issues. Unsurprisingly, the more that people have used AI, the more likely they are to report benefits in things like my speed at getting things done and my understanding of complex ideas and concepts. Again, indicating that if we can get people to use these tools, it may change their perceptions of them. Another thing that becomes clear with this study though, is that it's not just AI generally, but also the way that companies and people are interacting with AI that is causing issues. When asked which potential impact of generative AI on society is more likely, that business leaders are fully honest about job cuts, or that business leaders aren't fully honest with employees about job cuts, unsurprisingly, 7 in 10 folks in the US said that business leaders aren't being fully honest with employees about job cuts, which certainly is feeding into the anti-AI narratives. And something that I absolutely berate companies that pay me to come talk to them about. When people were asked what would increase their enthusiasm for using generative AI in work and life, two answers relating to employers scored highly. 57% of US respondents said that their enthusiasm would be increased if they were getting high-quality training through their employer about how to use AI effectively, and 59% said that their enthusiasm would increase if they felt sure their employer was using AI to increase productivity vs. eliminate jobs. One of the things that I talk about all the time with any company who will listen is that you have to have an open and honest conversation with your employees about how your leadership is thinking about AI. That does not mean that you have to pretend that there's no situation in which changes in the technology landscape aren't going to impact certain roles and jobs. But to the extent that your company views AI as an opportunity creation technology, not just an efficiency and cost-cutting technology, the more you can do to articulate that and be real with it, the better off you're going to be and the more employee buy-in you're going to have. They also found that long-term job security boosts likelihood to embrace AI. Among those who said that their job security due to AI was increasing, 50% said that they embraced AI as opposed to just 21% who said their job security due to AI is decreasing. And by the way, for those who think that this is a partisan issue, it is actually wildly nonpartisan. Going back to that question of what would increase enthusiasm for using generative AI, among workforce priorities, we heard about that high-quality training, but they also asked about the idea of employers being required to retrain or redeploy employees that were displaced by GenAI, and very similar percentages of left, center and right folks said that those things would increase their enthusiasm for GenAI. On the training question, 60% of the left said it would increase their enthusiasm, 61% of the center and 67% of the right. On the retraining requirement, you might think that the right, who historically have antipathy towards markets being forced to do anything, would be the lowest, but they're actually the highest, once again, at 60% as compared to the left's 59% and the center's 54%. But surely when it comes to government priorities like safety nets, we're going to see more of a divide, right? Not according to this study. When asked if an income safety net for those who lost their jobs to GenAI would increase their enthusiasm, the center was the lowest at 57%, then the right at 59% and then the left at 63%. All very similar numbers. And around government programs supporting the use of GenAI, once again, the right was highest at 60% increasing their enthusiasm, as opposed to 57% for the left and 54 for the center. Getting at part of why I think some people are having a hard time with just the barrage of AI in every part of their life. Edelman concludes that people who distrust AI are more likely to say that AI is imposed on them. In the US, 48% of people who trust AI said that they feel that generative AI is being forced upon them whether they want it or not. And that jumps to 67% when it comes to those who distrust AI. Now as we move into the remedy section, it's clear that the pathway to changing this is not going to be through business leaders or government leaders or probably even AI researchers. Instead it's going to have to come from our peers. When asked how much they trust different groups to tell the truth about generative AI, in the US government leaders came in even lower than CEOs, 24% compared to 27%. AI researchers were at 53%, still significantly lower than friends and family who are at 71%.