Cakir: Full Automation Should Not Be Used With Large Language Models
Tim Cakir · He Let Go of 50 People After One Tool Changed Everything · Jan 20, 2026 · at 30:01
AI trainer Tim Cakir discusses AI implementation strategy with host Sean Weisbrot, arguing that LLM processes should remain semi-automated with human approval.
0:00 / 0:07exact quote · 7.8s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“And I say semi-automate because I still don't believe in like when AI, when large language models are involved, I'm not into full automation yet.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Tim Cakir
Cakir let go 50 people after automating manual scraping with Clay
“I've kind of let go about 50 people. And that's because we got a tool called clay, you know and suddenly we didn't have to do manual scraping and we didn't have to do manual copy paste from LinkedIn and so on.”
Cakir: Companies need external AI trainers because top-down mandates spark panic
“You got to get a third party trainer. You got to get somebody to come in and help you train your whole team. Because when it comes from the leader, it's like, oh shit, you know, my boss is telling me to use it. If I don't use it, I'm in trouble. If I use it we…”
Cakir: AI is bigger than electricity and potentially fire
“So I think that AI is much bigger than electricity, potentially bigger than fire. And we haven't even understood it yet.”
Cakir replaced traditional CRMs with local Markdown files and Claude
“I don't have a CRM anymore. My CRM is Markdown files on my computer, right?”
Cakir: Leaders amplify AI fear internally due to alarming media coverage
“I think that the leader is in fear, first of all, because of the news and what's happening outside. So they fear it and they're trying, like, they're kind of amplifying that fear internally. A lot of leaders, I'm not going to say all leaders, but a lot of them…”
Cakir: AI labs market through fear because negative news amplifies faster
“Good news positive news doesn't amplify. It doesn't travel very fast. But negative news and scary news, news full of fear travels very fast and it amplifies very fast. So how do you market yourself is by saying a hundred million jobs are going to get lost to m…”