Technology
Air Canada Has to Honor a Refund Policy its Chatbot Made Up
O Canada. This is one of my favorite stories of the week. A Chatbot makes up some random stuff and the company has to pay up. Beware the chatbot powered by LLMs. Of course, a human could just as easily have mis-coded a rule in a deterministic system to yield the same result, it’s just a lot more fun when a computer does it.
Technology
3 Most Common Problems with Small Language Models
The are a variety of acronyms currently raging through the AI landscape—RAG, SSM, SLM. SLMs, small language models, are of interest because they are more efficient (lower compute cost, faster training) and are more easily fine-tuned for specific tasks. Like most short cuts, it’s often easier said than done.
Marketing
How Brands Can Embrace AI Hallucinations
With AI tools purported to hallucinate anywhere from 3-27% of the time, maybe we should be encouraging this behavior, especially if you’re a brand marketer
My favorite quote:
“LLMs [large language models] are bad at everything we expect computers to be good at,” he says. “And LLMs are good at everything we expect computers to be bad at.”
Operations
Gartner Predicts that with Advances in GenAI 30% of Fortune 500 Companies Will Consolidate Postpurchase Tasks Into a Unified Customer-Facing Role by 2027
By 2028, the EU will mandate “the right to talk to a human” in customer service interactions.
I’m not sure this is a good thing. Humans are so hit or miss. Sure, they might hallucinate less than AI (how would we know) but are they better than AI? I think the jury is still out and that’s after a billion years of evolution. Computers have been at it for only a couple of decades.
Shameless
B2B Wins #24: From Mad Scientist's Workshop to Well-Oiled Machine: Governance for Scaling Startups
This week, I got on my soapbox to deliver a framework for better managing your business. This was primarily focused on helping startups but applies to anyone managing an organization. It answers the age-old question: How do I align this team and get them laser-focused on our goals?