Everyone is into generative AI as of late. We all endure from ChatGPT syndrome. ChatGPT entered into our lives and now nothing will ever be the identical. Scammers and entrepreneurs alike, college students and entrepreneurs, tech bro influencers and YouTubers, everybody has one thing to say about generative AI. Everyone is a immediate engineer now. The hype is robust. Maybe lower than ten months in the past, however nonetheless…
I rode the hype like a napoleonic steed, demo after demo, tutorial after tutorial. Here, I’ll attempt to current what I discovered within the course of.
First of all, is it me or have we exhausted all of the doable demo apps we may think about ? I did quite a lot of them. And now, once I take a look at my AI-dominated Twitter feed (sorry not sorry Elon), I can’t assist however discover the identical demos or barely completely different variations of the identical demos over and over. With ChatGPT, OpenAI opened Pandora’s field, and it’s empty now, it appears. Here are the core concepts you will notice repeatedly:
- RAG like demos: Chat with PDF/ web site content material / YouTube video / no matter. I’m responsible right here. I lately constructed Discute to assist folks chat with their information base.
- AI writing instruments: Summarise paperwork, generate essays, articles, advertising copy because of LLMs. Here I severely have no idea why somebody would pay for an extra answer when ChatGPT or Bard or no matter proprietary LLM chat interface can already allow you to do this. I assume the important thing right here is advertising and laziness. With sufficient advertising spend, you’ll be able to basically persuade folks to pay to your product to get a barely extra handy expertise (a bunch of kinds for customers to state their preferences) than ChatGPT, seize a pleasant a part of the market and simply journey alongside until folks notice how low high quality the articles generated are and churn or keep (if they don’t care about high quality).
- AI element to an present kind of product: Here I put all of the AI-powered issues. AI-powered customer support instrument, AI-powered HR instrument, and so on. The downside with these options is that they don’t have any moat since each incumbent of their market can simply add these sprinkles of AI to their product, and so they have already found out the distribution. In the quick to medium time period, with sufficient advertising funding because of VC backing, you’ll be able to most likely seize sufficient customers to really feel like your thought…
