When employees at Meta learned that the company would start collecting their data to train AI models, there was a swift backlash. Many of them openly pushed back on Meta’s plan to capture keystrokes and mouse movements, circulating an online petition …
Meta may have been a bit too enthusiastic in its highly publicized campaign to reduce middle management.
The social media giant has reportedly been asking some individual contributors in its Applied AI (AAI) division if they would like t…
One of my clients came to me with a high-stakes pitch to request a budget and head count for her team. She had built a beautiful and thorough deck with revenue charts, growth graphs, staffing history, and supporting data. Buried somewhere behind all t…
A quiet but meaningful shift is happening in how people communicate—and most organizations don’t yet recognize its health and business implications.
Over the past 15 years, research shows that people are speaking hundreds of fewer words per day, re…
When ChatGPT was first released in 2022, I, like many people, gave it a try. I was mesmerized by how quickly it could generate a poem about my morning bagel. However, after giving it a few work-related tasks, I abandoned it. Not for me, I thought, finding the output too shallow or unreliable.
Fast-forward almost six years, and AI is now deeply embedded in the workflows of my solo business. I’ve learned how to work with the technology and guide it to the result I want.
Back in 2022, when I decided that AI didn’t work for me, the real problem was the way I’d set it up. Yes, the models have gotten substantially better, but AI frustration often comes from not knowing how to work with the tools. It’s understandable, since AI doesn’t come with an explicit “how-to” guide the way other software does.
The advantage of running a business by yourself is that you can try new AI tools, run a few tests, and change your setup quickly. Here are some common problems I’ve run into, and how I’ve solved them.
Problem 1: Not having a clear use case for AI
Census Bureau data shows that companies with four or fewer employees have an AI adoption rate of 20%. Compare that with 37% of firms with at least 250 employees.
Certainly, large companies are more complex and may have a greater need for AI. But they also have more resources. They can run 10 pilot programs and write off 9 of them. You may try one or two scenarios, without good results, and feel like you’ve wasted your time.
Regardless of size, every company or solo business has to start with the same question: How exactly can AI benefit me in this specific scenario? An approach of “I should be using AI for something” won’t work. It’s too open-ended, and you don’t have a clear idea of what you’re looking for.
I start by asking myself, What am I doing that’s repetitive? A task you do every week and dislike is a great place to start. Then get specific about where AI fits into the task.
For example, let’s say you want AI to help you write social media posts for LinkedIn. “Write social media posts” is not a good prompt because you’ll end up with generic AI slop. But dictating your ideas, letting AI format the dictation into social media posts, and then letting AI send the drafts to a social media scheduling tool is a genuine time-saver.
Problem 2: Using a single AI tool for every task
Many solopreneurs start with a general-purpose chatbot (ChatGPT, Claude, Gemini). When the output is mediocre, the technology gets the blame. But sometimes it’s just not the right tool for the task. Useful tasks are often micro use cases: AI aimed at one narrow problem, and only that problem.
I use an app called Airtable to manage my business, and I constantly need to update information. If I add a new client, I log the client’s address. Now I can just add my client’s website, and Airtable AI can fetch the client’s address for me.
Plenty of software you already pay for has AI built into it, and that’s often the better option. It’s designed with that product’s data and that product’s workflow. Often, this is a onetime setup: You add the AI feature to a specific task, and then it runs in the background.
Problem 3: Lack of context
General-purpose chatbots can still play a role in your business, but many people don’t take the time to give the AI tool the right context. Think of it as onboarding a contractor (without implying that an AI tool is anything close to human). Whenever you hand over work, you need to provide specifics about the job that needs to be done and what you’re expecting to receive at the end of the task.
Every project I have set up in Claude comes with lengthy documentation about who I am, what I do, and how I expect Claude to help. Without it, the output comes back generic, and generic output is exactly what makes people decide AI can’t do a task.
That means I’m putting in time up front, sometimes a lot of time. The payoff later is real in reclaimed hours, but the setup is what some solopreneurs skip. They want to prove that the task is worth handing over to AI, so they skip straight to prompting and review the result. That never works without the up-front context.
Continue to review and refine your AI processes
A cofounder of Clay shared the software company’s internal AI policy on LinkedIn. The first guiding principle is that employees using AI must stand by every idea and sentence in the AI output.
AI is not autonomous, no matter how good the output looks. With anything that AI generates for me, I make the final call on its accuracy and usefulness.
The other half is refinement. If the output comes back wrong the same way every time, the fix is usually more context or clearer instructions. The more of your own judgment you build into the setup, the more useful and time-saving AI will be.
This article is written by Tina Hamilton, an Entrepreneurs’ Organization (EO) member in Philadelphia and the founder and CEO of myHR Partner, a human resources outsourcing firm that manages HR for small and midsize businesses throughout the United Sta…
Although leadership is often portrayed as a glamorous career destination, being put in a position of power or responsibility is by no means an indicator of your ability to actually lead effectively. In fact, decades of research highlight an uncomfortable and pervasive gap between the leaders we have, and those we actually need. Indeed, some of the traits that help people emerge as leaders appear far less useful once they actually have to lead: narcissism, for example, predicts leadership emergence but not leadership effectiveness.
Consider that the failure rate for corporate managers is estimated at least 50%, the confidence in our leaders continues to decline and is now at an all time low, and that the correlation between money spent on making our leaders better and people’s faith in our leaders is negative. Add to this the recurrent data on employee disengagement (caused mostly by their direct supervisors), and the fact that people join companies but quit their bosses, and the question seems rather unavoidable: if only a minority of leaders are capable of performing to a high level, then should fewer people try to become leaders? In other words, if a high-proportion of leaders become ineffective, incompetent or inept at their job, should fewer people consider leadership as a career option?
Conversely, we rarely talk about the importance of good followers, but they are actually a critical aspect and underrated ingredient of leadership excellence. First, while leaders should be evaluated through the performance of their team, teams will perform better if they have smart and aligned followers, which is why in most professional sports teams there is a strong positive correlation between a team’s compensation and their performance (which leaves coaches to account for a small percentage of the variance). Second, as all political elections show, democratically chosen candidates are always as good as their voters, so followers’ maturity, educational attainment, and social intelligence are all critical factors determining the result of elections — if you want smart leaders, it is helpful to have smart followers. Third, follower maturity (as well as competence and integrity) is key to keeping bad leaders in check, which is why incompetent, toxic, and destructive leaders are far more likely to remain in charge when their followers and subordinates are blind to their performance.
Which is it?
So, is it better to be a good follower or a bad leader? The answer is not straightforward. Polite etiquette and a prosocial impression would make most people answer with the former, which does not however stop them from preferring the status and success associated with the latter. After all, bad leaders are often quite good at retaining power, especially if they report to, well, other bad leaders. Furthermore, the most common reward for being a good follower is a leadership job, which is odd if you think about it: why would we stop individual contributors from doing a job they are good at to put them in a role they may not only be unable to do, but also disinterested in? “Thank you for being our star performer, you will now have to manage a team of fifteen… and stop doing what you are good at”. This explains the timeless popularity and relevance of the famous Peter principle, which states that sooner or later most people are promoted to their own level of incompetence. Indeed, whatever role you stagnate in is one step beyond the role you should have stayed in, which is to say that your potential is probably the maximum role you reach minus one.
On the other hand, the question will depend on whether the outcome is judged on the basis of your own or other people’s interests. To state the obvious, being a good follower is good for others but not so good for you if it limits your career success, which is what not becoming a leader (sadly) means, at least in corporate settings. Conversely, being a bad leader may still be good for you if it means attaining status and prestige, as well as higher levels of income and success, but this will come to the detriment of others, especially those who report to you.
A better approach
There is, however, a more useful way to think about this distinction, which is to stop treating leadership and followership as opposite ends of a hierarchy. The best followers are not passive, obedient, or deferential. They are competent enough to know when their leaders are wrong, confident enough to challenge them, and sufficiently committed to the collective interest to put the team ahead of their own ego. In that sense, good followership requires many of the same qualities we supposedly want in leaders: judgment, integrity, self-awareness, courage, and the ability to work effectively with others. Indeed, organizations would probably improve if they spent a little less time teaching everyone how to lead, and a little more time teaching people how to contribute, cooperate, and occasionally disagree with those who do.
Likewise, one of the clearest indicators of leadership potential should be the ability to follow. People who cannot tolerate being managed, who regard every disagreement with their boss as evidence of incompetence, or who need to be the smartest and most important person in every room are unlikely to undergo a magical personality transformation when promoted. Give them power and you are simply removing some of the constraints that previously kept their narcissism, impulsivity, or delusions of grandeur in check. Leadership does not cure character flaws. Quite often, it gives them a larger budget.
Perhaps, then, the real problem is not that too few people want to be good followers, but that we have designed careers so that leadership is regarded as the ultimate proof of professional success. Until organizations create equivalent status, recognition, and financial rewards for exceptional individual contributors, ambitious people will continue to pursue leadership roles whether or not they have the talent, temperament, or desire to lead. We should hardly be surprised by the result. If the only way to get ahead is to become the boss, we will continue turning excellent followers into mediocre leaders, while rewarding mediocre leaders simply for remaining in charge. A good follower may contribute more value than a bad leader, but unfortunately value and status are not the same thing. Much of the history of management can be understood as our persistent failure to notice the difference.
Feeling overlooked in meetings or left out of key decisions can signal a troubling shift in your workplace standing. This article outlines 12 specific warning signs that your professional relevance may be slipping, along with practical strategies to r…
When employees dare to organize their workplace, a $1.7 billion industry of consultants and law firms springs into action. According to the Labor Department, something like 75% of employers enlist union avoidance consultants to quash union efforts. So…
In some places, the next battle in the war over remote work may be fought in legislatures, not HR departments.
The United Kingdom, Ireland, and New Zealand have already passed laws giving workers the right to request remote work and to have the req…
