Show HN: Companies use AI to take your calls. I built AI to make them for you

Original Article ↗ Hacker News Discussion ↗

Enthusiasm for the Concept & Use Cases

  • Many like the idea of offloading routine or unpleasant calls (reservations, checking availability, basic info, navigating long IVRs).
  • Several introverted users or phone-averse people explicitly say they’d pay for this.
  • People see value in:
    • Calling multiple vendors (restaurants, plumbers, salons) to compare availability or price.
    • Negotiating with cable/insurance/phone companies.
    • Scambaiting and tying up scam call centers.
    • Acting as a first-line agent that triages issues then hands off to a human.

Skepticism, Externalities & “Making Things Worse”

  • Strong concern that this will massively increase low-cost outbound calls, wasting time for 29/30 businesses who don’t “win” the job.
  • Some argue that when humans had to make all calls, effort naturally limited this behavior; AI removes that friction and creates negative externalities.
  • Fear that outbound AI will just move spam from inboxes onto the phone network and worsen robocall fatigue.
  • Ethical worry: small businesses and workers being forced to interact with bots instead of real customers.

User Experience & Edge Cases

  • Many say they only call when they have edge cases; they don’t trust an AI to handle complex verification, unusual billing, or level‑3 problems.
  • Some praise existing LLM-based answering services that accurately capture details and hand off a clean summary to humans.
  • Suggestions:
    • Faster, snappier turn-taking and fewer “polite” goodbye loops.
    • Simple agents whose only job is “get me to a human/manager as fast as possible.”
    • An API/JSON interface rather than browser-only.

Resistance to AI Voice Agents

  • A substantial subset says they will hang up if they detect a bot, or actively prefer companies that guarantee “human-only” support.
  • Others don’t mind bots, but only if they’re clearly better than current IVRs and don’t block escalation.

Privacy, Security, Legal & Trust Concerns

  • Strong criticism of the public “live feed” exposing names, partial SSNs, account and card digits.
  • Worries about giving sensitive personal data to a third-party agent that may be a future target for scammers.
  • One commenter cites U.S. robocall law (TCPA) and recent FCC moves around AI voices, questioning legality of automated outbound calls.

Broader Reflections & Future of Agents

  • Some envision a future of “agent-to-agent” interaction (customer agent ↔ business agent), possibly displacing websites and apps.
  • Others see this as wasteful “computers talking to computers in human language” when APIs could be more efficient.
  • Speculation about handshake protocols where both sides agree whether to use AI or humans, and gradual migration from voice to text-based agent interfaces.

Product Feedback & Practical Questions

  • Questions about pricing transparency and data privacy language on the site.
  • Interest in: open-sourcing or self-hosting, non‑US and multilingual support, and use with government agencies (IRS, DMV).
  • Some note that similar products already exist and ask what’s novel here.