The AI Pendulum Is Swinging Wildly. You Don’t Have to Swing With It.
The conversation around AI has changed dramatically, but neither extreme tells the whole story.
From excitement to suspicion
Not long ago, the message surrounding AI in business seemed remarkably clear: use it now, learn quickly, produce more and work faster, or risk being left behind.
AI was presented as the answer to almost every challenge a small business owner might face. It could write our content, create our images, research our ideas, communicate with customers and help us achieve in minutes what previously took hours.
For many people, that promise was exciting. For others, it created a new form of pressure. Even those who were uncertain about the technology were repeatedly told that adopting it was no longer a choice, but something every forward-thinking business owner simply had to do. If you did not adapt, you were told, you could be out of business within a year.
Now, the conversation sounds very different. We are hearing growing concerns about AI slop, misinformation, copied ideas, manufactured expertise and creative work produced without sufficient human thought. Writers and artists are being asked to prove that their work is genuinely theirs, while people are questioning whether the photographs, profiles and messages they encounter online are real.
Meanwhile, the platforms that encouraged us to create more with AI are introducing AI-content labels on Meta, a “Seems like AI slop” reporting option on LinkedIn and the new Facebook Verified badge to help restore trust in what we see. The pendulum appears to be swinging away from unquestioning excitement and towards suspicion, frustration and, in some cases, complete rejection.
At one end of the conversation, we are still being encouraged to automate more, publish more and use AI to increase our output. At the other, people are stepping away from tools and platforms because they no longer feel comfortable with what the online world is becoming.
For heart-led business owners, this can make it even harder to understand what is changing and decide what to do next. Are we falling behind if we do not embrace AI? Are we contributing to the problem if we use it? Do we need to produce more to remain visible, or would it be better to withdraw from the noise altogether?
These are not simple questions, and the answer is unlikely to be found at either extreme. The AI pendulum may be swinging wildly, but we do not have to swing with it. We can pause long enough to understand what is changing, think critically about what we are being told and decide for ourselves where AI belongs in our work.
When making things easier became producing more
When AI first entered the business conversation, one of its greatest promises was that it could make our work easier. It could help us get started when we did not know what to write, organise our thoughts and research, explore ideas and complete everyday tasks more quickly.
For small business owners, this opened up possibilities that might previously have been beyond our time, capacity or budget. It also promised to give us something many of us wanted even more: more time away from our businesses and a better balance between our work and the rest of our lives.
More recently, AI agents have extended that promise even further. Rather than helping with one task at a time, they can carry out a series of connected tasks, such as gathering information, updating documents or organising parts of a workflow, without us having to guide every step.
That promise was real. Used thoughtfully, AI can help someone untangle an idea, notice something they have missed or find a clearer way to explain what they already know. For a business owner with limited time or capacity, this can be incredibly helpful.
But somewhere along the way, making the work easier became an expectation that we should do much more of it. If AI could help us write one post more quickly, we could now write five. If one idea could become a blog, an email, a carousel, a video and several short posts, we were encouraged to create them all. If it became possible to appear on more platforms, the suggestion was that we should be on more platforms.
Instead of reducing the work involved, AI was increasingly being used to increase the amount of content we produced. This pressure was not entirely new. Social media had already encouraged us to post regularly, follow trends and remain visible. AI simply made it possible to produce content on a much larger scale.
A recent interview with Gary Vaynerchuk showed how far this culture of volume has gone. He compared publishing five social media posts a week with doing only three push-ups and said that he had distributed 397 pieces of content in one day. However, he has a large team turning longer pieces of content into posts for numerous platforms and accounts.
That is very different from a small business owner creating thoughtful content while also working with clients, developing offers, answering emails, managing their finances and living a life beyond their business. Most small business owners are not running content companies. Social media is one part of their work, not the whole of it.
The thinking behind creating so much content is easy to understand. Every post creates another opportunity to reach someone. However, there is no guarantee that producing more will lead to better results. When everyone can produce more, the online space becomes fuller, but it does not necessarily become more useful or interesting.
AI can help us turn one worthwhile idea into several useful pieces of content. But it can also make it easy to create a great deal of content before we have anything meaningful to say. Just because we can create more does not mean we need to.
For some businesses, creating a large amount of content may make sense. For others, the greater value of AI may be that it reduces the time spent creating content and leaves more time for clients, conversations, creativity and the work only they can do. Making the work easier does not have to mean filling the space we have saved with even more work.
When more content creates less trust
The internet was already noisy before generative AI arrived. Misleading information, exaggerated promises and copied ideas are not new. Nor is the tendency for people who sound confident to attract more attention than those who genuinely know what they are talking about. AI did not create these problems, but it has made it much easier to reproduce them quickly and on a far larger scale.
Someone can now publish polished, confident advice about a subject they barely understand. Images can depict events that never happened, while a single idea can be turned into dozens of posts within minutes, whether or not there is enough substance behind it. As the amount of content grows, it becomes increasingly difficult to know what we can trust.
We can see the effects of this in different ways. Fabricated images can be mistaken for evidence of events that never happened. At the same time, knowledgeable people are questioning whether they still want to share their work online.
A recent Google Earth experiment demonstrated how quickly this can become a much larger problem. Google introduced an AI feature that allowed users to generate fictional scenes using its satellite, aerial and 3D imagery, but withdrew it only hours later after people created images appearing to show disasters, attacks and other events that had never occurred. Although the images were watermarked and did not alter the public map, screenshots could still be separated from that context and circulated as apparent evidence. The incident showed that the ability to create an image is developing faster than the judgement and safeguards needed to decide where that ability belongs.
The loss of trust is not only affecting what people believe. It is also affecting whether those with genuine knowledge still want to share it online.
A recent Instagram carousel from an experienced acupuncturist brought this problem to life. She had decided to step away from the platform because she no longer wanted to compete in an environment where confidence was rewarded more than competence, copied content gained attention and responsible information was overshadowed by dramatic claims.
At one end, businesses are being encouraged to create hundreds of pieces of content. At the other, experienced people are withdrawing because they no longer want to participate in the volume, exaggeration and pressure. For heart-led business owners, the choice can begin to look like becoming a content operation or disappearing completely.
But those are not the only choices.
We can use AI without allowing it to manufacture our expertise. We can communicate clearly without exaggerating our claims and share something because it is useful and worth saying, rather than because another space in the content calendar needs to be filled.
AI has increased the amount of content we can create, but the more important question is whether what we create gives people a reason to trust us.
When the platforms begin rebuilding trust
The platforms themselves are now responding to this loss of trust. On LinkedIn, users can report a post because it “seems like AI slop”. Meta places an “AI info” label on some content when it detects signs that AI was used or when the person publishing it has disclosed this. Facebook is also introducing Facebook Verified, a badge intended to show that the person behind a profile has confirmed their identity.
There is an interesting contradiction here. The same platforms that encouraged us to embrace AI, create more and work faster are now introducing reporting options, labels and verification systems to help us trust the online world this increase in production has created.
But these measures can only tell us so much. A badge may confirm that a real person is behind a profile, but it cannot tell us whether that person is knowledgeable or trustworthy. A post may have been written without AI and still contain little original thought, while another may have been developed with AI and remain firmly rooted in someone’s experience and expertise.
The real question is not simply whether AI was used. We also need to consider how it was used, how much human thought shaped the finished work and whether the person publishing it is willing to take responsibility for it.
Trust cannot be restored by a label alone. It grows over time, when people can recognise the knowledge, judgement and human presence behind what they see.
When suspicion is mistaken for evidence
As concern about AI grows, the pendulum can also swing too far in the opposite direction. A recent publishing dispute shows how difficult this is becoming.
A highly anticipated novel by Jerry Falade attracted an offer of more than $2 million before his agents withdrew the manuscript from sale, saying that they could no longer authenticate how it had developed. Falade strongly denies using AI, and the full circumstances remain disputed, as reported by The Guardian.
The story raises an important question. How do we judge whether something has been created by a person, with AI or through some combination of the two?
It is right to question work that may have been produced dishonestly or without enough human involvement, but suspicion is not the same as proof. Human writing can sometimes be identified as AI-generated simply because it is clear, structured or follows familiar patterns. AI learned from human language, so some of the qualities we now associate with AI were present in human writing first.
Critical thinking is needed in both directions. We should question how AI has been used, but we should also question the evidence being used to judge it. The thoughtful middle is not only about using AI responsibly. It is also about judging its use responsible.
Choosing where AI belongs in your business
AI can help us organise research, explore ideas, notice gaps and communicate what we already know more clearly. The important distinction is whether AI is supporting our thinking or doing the thinking for us.
Thoughtful AI use means remaining involved. We provide the ideas, question the output, check the facts and decide what belongs.
When creating or assessing a piece of work, we might ask:
- Where did the original idea come from?
- How much human knowledge and thought shaped the result?
- Has the information been checked?
- Does the finished work genuinely reflect the person publishing it?
- Are they willing to take responsibility for it?
There is no single correct way for every business owner to use AI. We do not have to adopt every new tool to keep up, nor do we have to reject AI completely to protect the human qualities in our work.
We can decide which tasks we are comfortable asking AI to support and which we want to keep entirely in our own hands. We can also use it to reduce our workload without automatically filling the time we have saved by producing even more.
For one person, this might mean using AI extensively behind the scenes while remaining closely involved in everything they publish. Another may use it only for research or routine tasks. Someone else may decide that particular tools or uses do not feel right for them at all.
We can also change our minds as the technology develops and we understand its effects more clearly. A thoughtful position does not need to sit exactly in the centre. It simply needs to be chosen with awareness, rather than dictated by excitement, fear or pressure.
The important thing is that the choice remains ours.
Let the pendulum swing without taking you with it
The conversation around AI will continue to change. New tools will appear, new concerns will be raised and there will be more promises, controversies and pressure to decide which side we are on.
We do not have to react to every swing. We can remain curious without being swept up in the excitement, and cautious without rejecting everything AI might help us do. We can pay attention to the risks while still recognising the possibilities. Most importantly, we can continue to think for ourselves.
AI does not have to replace our creativity, knowledge or judgement. Nor does it have to become another demand to work faster, publish more and be everywhere at once. It can simply be a tool we choose to use where it genuinely supports us.
The AI pendulum may continue to swing wildly for some time, but we do not have to swing with it. We can pause, notice what is changing and choose a relationship with AI that allows us to remain recognisably present in our own work.