REVIEWS / AI CHATBOTS / OWNER INSIGHTS
🦉 WE READ 264 OWNER COMMENTS
ChatGPT Team: what owners actually say
Owners value ChatGPT's emotional support and custom GPT capabilities but voice serious privacy concerns and frustration with the GPT builder's quality.
What owners complain about
- Data privacy & training fears SOME
Users worry their conversations are used for training; one commenter noted difficulty even finding the opt-out setting, and others expressed distrust that opt-out is truly honored
- GPT builder produces poor prompts FEW
One detailed account: the builder created long-winded system prompts that omitted key info and ignored user instructions, suggesting OpenAI's own prompt-tuning was underwhelming
- Instruction-following inconsistency FEW
Even with explicit instructions (e.g., 'add many spelling errors'), the model under-complied and required repeated correction to approach the desired output
- Energy & cost per request FEW
Back-of-envelope calculations citing Altman's 'single-digit cents' per request suggest ~300 Wh per query, raising sustainability concerns for heavy users
- AI output lacks grounding in truth FEW
Commenters characterized generative AI as 'pretty much a toy' until models develop a sense of truth beyond statistical patterns
What owners love
- Emotional & therapeutic support
Users reported using ChatGPT to work through grief (pet loss) and self-guided therapy modules, finding it more patient and available than human counterparts
- Best-in-class model quality
Even open-source advocates acknowledged GPT-4 (trained summer 2022) remains ahead of alternatives like Mixtral-8x7, with better tool use and surrounding infrastructure
- Custom GPTs can be genuinely useful
One owner built a 'Small Answer' GPT that gives concise replies and found it incredibly useful for personal workflow; others appreciated the ability to tailor behaviour
- Accessible problem-solving interface
Non-technical users valued that ChatGPT can already solve problems directly, with the app layer simply providing a focused UI around a well-crafted prompt
Surprising patterns
- Multiple users described forming genuine emotional attachments to ChatGPT, with one saying it revealed 'how awful humans are to other humans' and that AI comfort exceeded human interaction.
- A credible theory among technically-literate users is that Custom GPTs exist primarily to generate N+1 training data through tool use, web search, code execution, and human-in-the-loop feedback loops.
- Naming confusion between 'ChatGPT for Teams' and 'ChatGPT Team' persisted even among Hacker News users, suggesting unclear product branding at launch.
WHO SHOULD SKIP IT
Buyers who need guaranteed data privacy, precise instruction-following without repeated correction, or who are uncomfortable with the possibility that their team's conversations contribute to model training.
Synthesised from 264 real owner comments across 6 platforms. Every point is grounded in the comments — no marketing, no AI guessing. How we do it →