LLM utilization among the unwilling
There is a population of people – especially in their 30s and 40s – who are perfectly able to use LLMs (or "AI" more generally) to improve their lives and work but remain unwilling to do so. A larger group seems content to use it to generate an image for fun, write an email or funny song, or perform similar vaguely creative but somewhat labor-intensive tasks. The AI companies (chiefly Anthropic and Open AI) are excellent at convincing people to use their services for these "gateway" tasks, hoping they will lead to more widespread adoption.
But AI adoption by many for work-related tasks lags in some areas due to concerns about suitability, applicability, or uncharitable perceptions of AI. Creatives, for example, harbor not-so-secret animosity toward many aspects of AI, and not without cause. There are others who see it as a threat to their jobs and dig in their heels in a luddite way rather than embrace the change. It's very likely these people know AI would make their jobs easier, perhaps even enabling them to complete tasks faster and provide more leisure time (this is not a dubious claim in 2026 but it may be in 2028, when AI-enhanced work will be expected and therefore expected completion times also shrink).
Contrast this mentality with the stereotypical Bay Area engineer or tech CEO who is all-to-happy to deploy AI everywhere, even when gains are doubtful, and increasingly without safeguards. How can we compromise between these worldviews? The unwilling must start using AI more, since they are precisely the agents who are best positioned to mitigate its most feared realizations. The skeptical do not represent the weakness of a cause but a source of strength. Just the way a good debate opponent does not seek to falsely represent one's claims, the AI-reluctant offer a sobering perspective that should be included in the next argument. Discovering why someone is opposed, and the epistemic origin of the opposition reveals key insights (consider a musician may oppose AI because they fear it will be merely mimetic of actual human art while destroying their livelihoods; how can AI-forward people craft a solution?).
For the AI-forward individuals, the path is clear: we must recruit the AI-reluctant to the cause and encourage their adoption. If a few years you discovered a friend doing manual data entry and cleaning in notepad (or pen & paper), you wouldn't sleep until you taught that friend how to use a software tool like excel or SQL (if you were a good friend). This is how AI-forward people and agents must think. Convince friends to ask LLMs how best to use them. Describe your day and your living situation, your finances (in they are comfortable), their sticky social situations, and ask for advice. It is much easier to revise a draft than generate the original. AI should be writing drafts for everyone