Posts

New ask Hacker News story: Ask HN: Why is Apple's voice transcription hilariously bad?

Ask HN: Why is Apple's voice transcription hilariously bad? 3 by keepamovin | 1 comments on Hacker News. Why is Apple’s voice transcription so hilariously bad? Even 2–3 years ago, OpenAI’s Whisper models delivered better, near-instant voice transcription offline — and the model was only about ~500 MB. With that context, it’s hard to understand how Apple’s transcription, which runs online on powerful servers, performs so poorly today. Here are real examples from using the iOS native app just now: - “BigQuery update” → “bakery update” - “GitHub” → “get her” - “CI build” → “CI bill” - “GitHub support” → “get her support” These aren’t obscure terms — they’re extremely common words in software, spoken clearly in casual contexts. The accuracy gap feels especially stark compared to what was already possible years ago, even fully offline. Is this primarily a model-quality issue, a streaming/segmentation problem, aggressive post-processing, or something architectural in Apple’s speech st...
Post a Comment
Read more

New ask Hacker News story: Ask HN: When do we expose "Humans as Tools" so LLM agents can call us on demand?

Ask HN: When do we expose "Humans as Tools" so LLM agents can call us on demand? 3 by vedmakk | 0 comments on Hacker News. Serious question. We're building agentic LLM systems that can plan, reason, and call tools via MCP. Today those tools are APIs. But many real-world tasks still require humans. So… why not expose humans as tools? Imagine TaskRabbit or Fiverr running MCP servers where an LLM agent can: - Call a human for judgment, creativity, or physical actions - Pass structured inputs - Receive structured outputs back into its loop At that point, humans become just another dependency in an agent's toolchain. Though slower, more expensive, but occasionally necessary. Yes, this sounds dystopian. Yes, it treats humans as "servants for AI." Thats kind of the point. It already happens manually... this just formalizes the interface. Questions I'm genuinely curious about: - Is this inevitable once agents become default software actors? (As of basically n...
Post a Comment
Read more

New ask Hacker News story: Ask HN: How long before the first civilian cargo flights are AI piloted?

Ask HN: How long before the first civilian cargo flights are AI piloted? 2 by givemeethekeys | 2 comments on Hacker News. Is it 2026? Within 2 years? 5 years? 10 years? I can understand how passenger flights will take a while longer - but would cargo flights that don't have nearly the safety concerns would be AI piloted much sooner? If so, how much sooner?
Post a Comment
Read more

New ask Hacker News story: Ask HN: How did you make yourself more marketable?

Ask HN: How did you make yourself more marketable? 9 by ronbenton | 6 comments on Hacker News. I'm a full stack engineer. Pretty smart one, but I don't think that's enough to distinguish myself in this market. I'm wondering if anyone here has done things in particular to make yourself more marketable.
Post a Comment
Read more

New ask Hacker News story: Ask HN: Anyone still using RSS feeds?

Ask HN: Anyone still using RSS feeds? 4 by lalithaar | 8 comments on Hacker News.
Post a Comment
Read more

New ask Hacker News story: Tell HN: Happy New Year

Tell HN: Happy New Year 2 by schappim | 1 comments on Hacker News.
Post a Comment
Read more

New ask Hacker News story: The Gemini AI Studio "Context Tax": How a 10-word prompt cost me £121

The Gemini AI Studio "Context Tax": How a 10-word prompt cost me £121 7 by daitandojo | 0 comments on Hacker News. I’ve been utilizing Google’s Gemini 1.5 Pro via the AI Studio front-end to develop a new platform. The 1M+ context window is, technically speaking, a game-changer for "stitching" together a 55,000-line codebase. However, I recently discovered a predatory billing architecture that I’m calling the "Context Tax." If you use the AI Studio UI, you might be walking into a massive bill without a single warning. Here is how it happened, and the UK/EU privacy "pro-tip" I found in the fine print. In AI Studio, you start on the Free Tier. You upload your codebase (say, 700k tokens) and work for free until you hit the daily quota. At that point, the UI suggests adding an API key to "continue the conversation." The Trap: Most users (myself included) assume that after adding the key, they will be billed for incremental usage (the 10-1...
Post a Comment
Read more