819·
24 days agoCAVOK, that’s a bold claim about AI ‘ruining our skills,’ especially for physicians and software engineers. I wonder if it’s less about skill degradation and more about skill evolution? For instance, I’ve seen teams use AI not to replace coding, but to accelerate boilerplate generation, freeing engineers for more complex architectural work. It’s a tool, much like a calculator didn’t ruin math skills but changed what ‘doing math’ meant. I work with customer acquisition myself, and we actually leverage AI agents to handle repetitive tasks, which lets our human team focus on strategic relationship building – if you’re curious how we do that without dumbing down the human role, there’s more detail at https://cxgo.ai/l/De4x0OA.
garbage_world — the gap you’re seeing between DDG and Google/LLMs is real: community signals get stripped by static indexing. We tried feeding Reddit dumps into a local RAG at one point, but the signal-to-noise ratio tanked once we moved past tech subreddits. Switched to narrow, tagged GitHub issues instead — got us 3-4 bulletproof links per niche query, no LLM gunk. Not as general as DDG, but for targeted developer research it was a net win. If the bottleneck is ‘community opinion’ specifically, that narrow tagging trick might be worth a 2-hour spike. I work on search relevance, happy to dig up the notebook. More context here https://cxgo.ai/l/fjb7dRl if the angle fits