The models behind GPT chatbots will continue to improve, but they are good enough today. Everything we discussed in the previous sections (1, 2, 3, 4) is possible with the technology that exists today. Model capabilities and personality do play a role in retaining users on a given platform. Trying to win the fight for…
AI Agents are a broad and often confusing topic. As we continue thinking about monetizing consumer GPT chatbots, let’s focus on AI Agents’ ability to take actions. Many people have seen agents that operate like a human using a computer or a browser. These are a useful demonstration of what will be possible to automate,…
The conversational nature of GenAI chat lends itself to voice modalities much more naturally than web search engines do. Speaking a keyword query and having results URLs read back to you feels very stilted. Search is a discrete action and doesn’t carry state from one search to the next. A voice conversation with a search…
Advertising is coming for GenAI consumer chatbots. In part 1, we looked at the parallels between web search and information retrieval with chatbots. Chatbots already have a rich collection of user data that can be leverages for ad targeting. Let’s take a look at how they can build an advertising platform to monetize usage. The…
Consumer-facing GPT chatbots will inevitably start to monetize their products. Cost pressures from burning through cash and the high costs of inference may accelerate this change. But even without the cost pressure, it seems unlikely that tech companies will leave any potential profits on the table. The standard game plan for tech companies providing “free”…
AI Agents are often characterized as smart workers who learn and improve with practice. This can be true, but it isn’t automatic, and the “learning” doesn’t happen in the way I see a lot of people assume it does. Let’s start with what makes an AI Agent: 🌍An understanding of the world (Large Language Model)…
Lemme explain. There is no feature or any different behavior when GenAI “hallucinates”. GenAI is always using probability to predict the next token to output. We just choose to call it a hallucination when it does something we don’t want. Why is this distinction important? I heard a training explaining hallucination as I did above,…
There are two categories of AI systems: Symbolic AI and Deep Learning. Symbolic AI requires the builders to define the rules it will apply. If you are trying to understand language, you would need to add grammatical rules. Deep Learning, on the other hand, starts with no rules and creates rules that match the patterns…
“The AI you use today is going to be the worst AI you ever use” is a phrase commonly shared by AI educators. This makes sense from a technical development perspective. Even if the scaling laws don’t hold, surely, we will find other ways to improve current genAI systems. While that is likely true, it…
For transformers and recommendation algorithms, it is all about attention. Social media platforms want to capture as much of your attention as possible in order to show you ads and collect metadata about you. LinkedIn recently notified users that it would utilize their content to train AI models unless they opt out. I’m curious if…