The AI Divide: Why Everyone Seems to Have Extreme Views

When developer Daniel Jalkut recently posted "My take on AI is, essentially, everybody who's against it is too against it and everybody who's for it is too for it," he struck a nerve that resonated across the Mastodon community. The simple observation about AI polarization sparked a conversation that revealed much about how we discuss artificial intelligence in 2026.

The post attracted a variety of responses that illustrated Jalkut's point precisely. Some users agreed enthusiastically with his middle-ground stance, while others pushed back from both extremes. This pattern of extreme positions on AI isn't new, but it has intensified as large language models have become more integrated into everyday development workflows.

Why AI Discourse Has Become So Polarized

Several factors contribute to the extreme views on AI that Jalkut identified. Jonathan Joelson offered an apt comparison: "It's like that bit 'everyone who drives slower than me is an idiot and everyone who driver faster than me is a maniac.'" This suggests that AI opinions may follow a similar pattern, where any position not matching one's own appears extreme.

Samuel Smith pointed out what might be sampling bias at play: "Those that are for/against are the most vocal. Those with nuanced opinions might be keeping them to themselves." This creates the illusion of polarization when more moderate positions may actually be common but less visible in public discourse.

The nature of AI itself contributes to this divide. As alfian noted, "It's impossible to be neutral towards LLMs. It feels like magic." This perception of AI as almost supernatural makes rational discussion challenging, as people's responses are colored by awe, fear, or both.

The "Too For It" Camp: Enthusiasts and Their Optimism

On one side of the divide are those who embrace AI with enthusiasm. Alfian's trajectory represents a common pattern: "I was unimpressed until 2026 rolled around and now I truly think it will change the nature of software engineering." Many developers who were initially skeptical have become converts as AI tools have improved dramatically in recent years.

This group often points to AI's ability to accelerate development, generate code that would take humans hours to write, and handle complex tasks that previously required specialized expertise. The productivity gains are undeniable, and proponents argue that AI will eventually augment rather than replace human developers.

Mx. Aria Stewart's simple "YEP THIS" in response to Jalkut suggests that many find the middle-ground position refreshing after years of extreme rhetoric. However, even this moderate view can draw criticism from those who believe any enthusiasm for AI is misplaced.

The "Too Against It" Camp: Critics and Their Concerns

On the other side are those who view AI with deep skepticism. Jacen's response captures this perspective: "im 'too against' boiling oceans. Centirst logic. Go figure." This suggests that some critics see any middle ground as equivalent to complicity with harmful technologies.

Carolyn brought up practical concerns with a pointed question: "Do you want a data centre in your back yard?" While initially unclear, her follow-up clarified the concern: "It could affect the balance." This represents a common criticism of AI - that the environmental costs and resource requirements are often ignored in enthusiastic discussions about its benefits.

Cedar Fen Farm took a more fundamental stance, arguing that "AI is a term no longer useful. It has become political and meaningless in a the hands of marketers and a public that have no idea what digital algorithms are." This perspective suggests that the term "AI" has become so overloaded that meaningful discussion has become nearly impossible.

The Middle Ground: Nuanced Positions and Their Challenges

Jalkut's original post represents an attempt at finding middle ground, but even this position faces criticism. As ran mak responded: "you're definitely being too in the middle!" This suggests that in the current discourse, any position not firmly aligned with one extreme or the other may be seen as insufficiently committed.

Collin Donnell expressed a common ambivalence: "I think it's useful but also kind of wish LLMs didn't exist?" This captures the complexity many developers feel - recognizing utility while having ethical concerns about the technology's existence and development.

Matt K expanded on this ambivalence: "I've witnessed it cranking out moderately decent code at a blistering pace but the consequences will be devastating. Cat is out of the bag now though so wishing won't help." This represents a pragmatic middle position that acknowledges both the benefits and risks of AI.

Practical Implications of AI Polarization

This polarization has real consequences for how AI is developed, deployed, and regulated. When discourse is dominated by extreme positions, nuanced discussions about appropriate safeguards, ethical guidelines, and responsible development become more difficult.

The divide also affects how individual developers approach AI tools. Some may feel pressured to either fully embrace or completely reject AI technologies, even if their personal views are more nuanced. This can lead to missed opportunities for thoughtful integration of AI into development workflows.

Additionally, the polarization makes it harder for the public to understand AI's capabilities and limitations. When experts themselves hold extreme positions, public understanding becomes even more fragmented, potentially leading to poorly informed policy decisions and public reactions.

Finding Balance in an Extreme Conversation

As AI continues to evolve and become more integrated into our technological infrastructure, the need for balanced, nuanced discussion becomes increasingly important. Jalkut's observation serves as a reminder that most complex technologies elicit a range of legitimate concerns and benefits.

The challenge lies in developing a discourse that acknowledges both the transformative potential of AI and its legitimate risks, without falling into the trap of extreme positions that shut down productive conversation. This may require creating spaces for more measured discussion, recognizing the validity of multiple perspectives, and moving beyond simplistic "for or against" frameworks.

As we continue to navigate the AI landscape, perhaps the most productive approach is not to choose sides, but to engage in the difficult work of holding multiple truths simultaneously - that AI can be both remarkable and concerning, both transformative and risky, both useful and potentially harmful. Only by embracing this complexity can we hope to develop and deploy AI technologies in ways that maximize benefits while minimizing harms.