Bluesky’s recent service hiccups caused a lot of buzz online, and it wasn’t just the outage that got people talking. Instead, a growing chorus of users took aim at the platform’s development methods, blaming so-called ‘vibe coding’ powered by AI for the disruptions.
The Outage and the Instant Reaction
On a Monday that saw several major platforms experience service problems, Bluesky was no exception. The social network, which has been gaining traction as an alternative to mainstream sites, suffered intermittent interruptions that left users frustrated. While the company quickly attributed the issues to an "upstream service provider," the response from Bluesky’s community was far from calm acceptance.
The moment the outages began, Bluesky feeds flooded with posts accusing the development team of relying on questionable AI-assisted coding practices, dubbed "vibe coding." People used this term, mixing casual slang with technical criticism, often with humor, irony, and sometimes real anger.
Users were quick to suggest that any bugs or failures stemmed from sloppy programming aided by AI tools rather than human expertise. Memes and sarcastic comments circulated widely, painting a picture of developers who had thrown caution to the wind in favour of quick, AI-generated code.
Decoding ‘Vibe Coding’
So what exactly is this so-called vibe coding? The phrase has become shorthand for a style of software development where coders lean heavily on AI assistants to generate or suggest code snippets, often without rigorous manual oversight. Critics argue that it leads to careless mistakes and a lack of accountability.
One vocal Bluesky user, T-Kay, captured the mood succinctly: “Any developer or programmer using ‘vibe-coding’ or any reliance on AI to code things is clearly too stupid to know how to do the job they’re paid to do and should be fired out of a cannon.” Harsh words, but they reflect a widespread mistrust of AI in the coding world among some users.
But the truth is more complicated. Professional developers in many corners of the tech industry are embracing AI coding tools, finding them useful for speeding up routine tasks and offering fresh ideas. But the backlash on Bluesky shows that end users often see these tools as a lazy shortcut, a scapegoat for any glitch or bug.
AI’s Growing Role in Software Development
Bluesky’s own team isn’t shy about their use of AI. Jay Graber, the founder and Chief Innovation Officer, openly acknowledged in late March that AI tools like Claude Code play a big part in their development process. Graber’s candour is a rare glimpse into a future where AI isn’t just a novelty but a fundamental part of building digital platforms.
Technical Advisor Jeromy Johnson, known on Bluesky as “Why,” has been particularly vocal about the shift. He revealed that Claude, an AI coding assistant, wrote nearly all his code in recent months. “Things are changing. Fast,” he said. Johnson’s admission shows a broader trend: AI isn't just assisting but sometimes leading the coding charge, reshaping how software is built.
AI promises efficiency and innovation, but it also raises questions about quality control, responsibility, and keeping the human touch in tech. Bluesky’s service troubles became a lightning rod for these anxieties.
Why the Resistance to AI in Coding?
The scepticism among Bluesky’s user base isn’t unique.
Many people remain wary of AI’s expanding role across industries, tech included. Part of the resistance stems from a fear that AI might replace skilled workers or degrade the quality of products by favouring quantity over craftsmanship.
But there’s also an emotional element. Coding has long been seen as a craft — a domain where human creativity, logic and expertise intertwine. The idea that a machine could shoulder much of this work, or that developers might rely on it too heavily, can feel like a betrayal of that craft.
There’s also the ‘black box’ problem: AI-generated code can be opaque, hard to fully understand or debug. That makes some users nervous, especially when things go wrong.
Still, the tech world keeps moving forward. AI coding assistants are becoming more sophisticated, with tools that integrate seamlessly into developers’ workflows, offering suggestions, automating tests and spotting errors before they become problems.
Looking Ahead: Balancing AI and Human Expertise
Bluesky’s situation is a snapshot of a larger conversation about AI’s role in software development. While AI can speed up coding and help catch mistakes, it isn’t a magic bullet. Developers still need to apply judgement, review AI’s output critically and maintain standards.
Sure, the social media storm over ‘vibe coding’ might be as much about trust as technology. Users want to know that the products they rely on are built with care and skill — not just spun out by algorithms.
So far, Bluesky’s leaders seem focused on mixing AI tools with hands-on human review. Whether that balance will satisfy a sceptical public we'll have to wait and see.
For now, the discussion is ongoing. And the question of what ‘good coding’ looks like in an age of AI is far from settled.
Related Articles
- Meta's AI push tests ad-era profitability
- Musk revisits 2015 rift with Larry Page at OpenAI trial
- AI demand forces memory makers to reroute capacity
As Bluesky’s developers press on with AI-assisted workflows, the backlash over ‘vibe coding’ shows that some users aren’t ready to embrace this new era without reservations. The tech community’s challenge will be proving that AI can enhance quality, not just speed — and that human skill remains at the heart of great software.
This article was created with AI assistance.
