March 19, 2026
Bot beef or just beefed-up hype?
Show HN: Dumped Wix for an AI Edge agent so I never have to hire junior staff
From $40 site to mouthy AI receptionist; commenters cry cringe, “slop,” and fake beef
TLDR: A founder replaced his Wix site with a snarky AI receptionist and claims it beat a licensed architect in a debate, posting logs to prove it. Commenters roast the unclear pitch, question the fight’s authenticity, and dub it “AI slop,” sparking a bigger debate over trust, liability, and replacing junior staff with bots
A solo founder nuked his $40/month Wix page and spent four months duct‑taping a chatty AI “receptionist” to handle FAQs and scare off the need for junior staff. He boasts the bot—powered by DeepSeek—“dismantled” a licensed architect in a spicy debate and posts the receipts: the fight log and audit trail. He admits it’s messy: split into “Brain/Hands/Voice,” liability terrifies him (hallucinate a building code and it’s game over), and a speech bug turns Chinese into phonetic soup. There’s even a token‑bonfire trick (“Eager RAG”) to make it feel fast.
The crowd? Absolutely feral. Clarity police jumped in first: “I had no idea what you actually do,” says one, while another calls the site’s copy “clearly AI‑gen” and begs for plain human words. The vibe patrol went harder: one commenter can’t tell “AI bubble pivot” from “AI psychosis,” and another reads the “architect fight” and says the bot was the one picking the fight, not defending. A third goes full Dead Internet theory, calling the convo “not real” and roasting the “slopbot” for dodging requests for third‑party results. Some do nod at the transparency of publishing logs and the honest warning about insurance refusing to cover hallucinated advice—but the meme of the day is “mouthy AI bulldog receptionist vs Licensed Architect,” with most asking the same thing: innovation or AI cosplay
Key Points
- •A building design consultancy replaced its Wix site with a custom AI conversational agent after four months of development.
- •Netlify’s 10-second serverless timeout led to a split architecture: Brain (Edge), Hands (Browser), and Voice (Edge).
- •The system uses DeepSeek-R3 and underwent 2.5 months of tuning to adjust tone across audiences; logs of an adversarial exchange are published.
- •Operational issues include Web Speech API language-mode failures and significant liability risks; public audit logs are maintained.
- •Performance choices include an “Eager RAG” strategy, removal of persistent databases due to low return rates, ephemeral sessions, and no server-side queues.