Great read from an opinion piece by an analyst on Fierce about AI and the hype. The main point that AI is really machine learning is interesting. The term “Artificial Intelligence” was coined in 1956. IBM Watson was launched in 2010 on Jeopardy! Amazon launched Alexa on Echo devices in 2014. So the current state of hype over AI is annoying.
The analyst suggests that AI is really just an LLM on top of ML. That Machine Learning has guardrails and AI really needs them.
The investment cycle and hype are Dot-Com-ish and there is no level of user adoption that will ROI the $109 Billion. “The US has bet overwhelmingly on the latter — pouring debt, subsidies and hype into systems that excel at generating plausible text but struggle with accuracy, reliability and ROI.”
We poured every bet into AI – and that has left other possible winners underfunded.
Isn’t ChatGPT similar to Alexa and Siri? Many people have found them less than useful. These voice assistants are early versions of ChatGPT and Claude. Claude and Manus are better wrappers for Alexa, but it all comes down to the prompting.
AI is inaccurate because it ate the whole internet — and a lot of the internet is bullsh!t, but it ate that too! So without really careful prompting and context, you have a big chance of getting a BS answer! “LLMs are probabilistic sequence predictors; they ingest information and take an educated guess at the most credible answer to a query.”
Agentic AI is hyped in CX, but how many people have been satisfied with their chatbot experience? Or their Voice AI bot experience? But maybe by design it was just a cheaper way for the brand to ingnore you.
A lot of work goes into the planning, design and implementation of CX and by extension AI in that department. Beyond the CX planning, there needs to be data – and that is where many of theese projects fail. Without domain knowledge (data), the AI is generic and unsatisfactory.
I think long term, Google and Microsoft win in the ChatGPT space due to size, eyeballs, data, and an understanding of monetization. CoPilot will have huge adoption because M365/Teams have huge adoption. It can be OpenAI under that wrapper or some other LLM. It won’t matter.
Google owns search. They will own this space too. They have too much at stake. They have too much money. They own the OS that runs more than 70% of mobile phones globally. (In the US, the Android market share is 40%). They already indexed the internet. Chrome has 55% of the web browser market. They just have to package it for profit.
Gemini will be ad supported by year end. You will have to watch a 20 sec commercial before you get your result.