Podcast takeaways: when automation should still hand off to a human
About our Podcast series: we quote public figures word for word from the linked sources, then translate the idea for Malaysian SMEs. They are not endorsing ThinkChat, and we don’t put words in anyone’s mouth.
Two of the loudest automation champions in business have, at different times, publicly walked part of it back. Their words are a useful guide for any SME putting AI on WhatsApp.
Elon Musk: “Humans are underrated”
In April 2018, while Tesla struggled to ramp up Model 3 production, Musk replied to a reporter on Twitter (now X):
“Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated.”
Elon Musk on Twitter, 13 April 2018, as reported by TechCrunch
In a CBS interview the same week, he described a “crazy, complex network of conveyor belts” that wasn’t working, so Tesla removed it.
Takeaway for CS: automating everything is not the goal. Automating the right things is. A factory found that out with conveyor belts; a business can find it out with a chatbot that won’t let customers reach a person.
Klarna’s CEO: “there will be always a human if you want”
Klarna’s co-founder and CEO Sebastian Siemiatkowski had been one of the most vocal advocates of AI-first customer service. In May 2025 he told Bloomberg the company was recruiting human support again. As quoted by Fortune:
“As cost unfortunately seems to have been a too predominant evaluation factor when organizing this, what you end up having is lower quality.”
Sebastian Siemiatkowski, to Bloomberg, quoted by Fortune (May 2025)
“[I]t’s so critical that you are clear to your customer that there will be always a human if you want.”
Sebastian Siemiatkowski, to Bloomberg, quoted by Fortune (May 2025)
A Klarna spokesperson also told Fortune the company was “very much still AI-first”. So this is not “AI failed”. It is a company learning where the human line sits.
A podcaster’s view of the bot
The same Fortune piece quotes tech writer and podcaster Gergely Orosz, who had tested Klarna’s bot. He found it “underwhelming”, but noted: “It recites exact docs and passes me on to human support fast.” He also praised its lack of made-up answers.
That is a pretty good description of a safe CS bot: sticks to the documents, doesn’t invent, gets you to a human quickly.
When automation should hand off
Put those lessons together and you get a practical rulebook for WhatsApp:
- Whenever the customer asks. “Can I talk to someone?” should never lead back to a menu.
- When the answer isn’t in approved knowledge, especially more than once in a row.
- When money or policy bends: refunds, deposits, discounts, exceptions.
- When emotion rises: complaints, frustration, urgency.
- When the stakes are high: medical, legal, financial questions, or a big order worth a human close.
And when it hands off, do it well: same chat, full context, honest wait time.
How ThinkChat builds this in
In ThinkChat, human takeover is available on every plan. When a customer asks for a person, the AI confirms, sends one short reply and pauses on that chat so your team can take over in the same inbox. If the AI can’t find an answer twice in a row, it can bring in your team automatically. It never makes up answers; it says “I don’t know” instead.
Automation does the volume. Humans keep the judgement. As Musk put it: humans are underrated.
Soft next step
If you want automation that always leaves a door open to your team, Talk to Lyia on WhatsApp: +60 11-5515 9318.
Plans stay simple: Junior RM299 · Senior RM599 · Manager RM899 a month. Pick the level that matches how much of your customer service you want covered.