Chapter 1: The Chatbot Phenomenon

Remember that scene from Arrested Development where Oscar Bluth launches ImOscar.com, always attaching "dot com" whenever someone mentions his name? Lately, the internet feels quite similar since the introduction of GPT-4.

The latest Duolingo Max is essentially Duolingo... with a chatbot! The updated Instacart resembles the previous version, but now it features a chatbot. Intercom has integrated a chatbot as well. Even Google has attempted to implement a timid chatbot (not built with GPT), merely shadowing advancements rather than leading them.

What did we think would happen? Following the debut of ChatGPT by OpenAI, which presented itself as a chatbot, a wave of companies began adopting this technology. It's akin to a person wielding a hammer — they tend to view everything as a nail. In my case, I would ponder how I ended up holding that hammer in the first place.

Businesses will adapt as necessary, which is understandable. If adding a chatbot allows them to maintain their core operations while providing extra value, why wouldn’t they? Generative AI also enhances their creative potential — at least in terms of quantity. The Duolingo chatbot appears beneficial, albeit at an additional cost beyond the standard subscription.

These are sustaining innovations: organizations are enhancing their current offerings for existing customers by leveraging new technologies. For years, Gillette introduced products with one extra blade at a time, until new competitors emerged with business models that swiftly gained market share.

Now, this discussion seems to lead us toward the disruptive innovations within generative AI. Disruptive innovations, such as Dollar Shave Club and its many copies, challenge the competitive edge of established companies (like Gillette) in ways that are difficult for them to counter, often requiring a complete overhaul of their core operations.

Gillette could have significantly reduced costs and product quality, eliminated third-party sales, and begun shipping $1 blades directly to customers through a subscription model, but they weren’t willing to make such drastic changes. This reluctance ultimately cost them dearly when customers favored the new business model.

I reference this example because it is archetypal. You are likely familiar with it, and that is precisely the point. I aim to evaluate the potential of generative AI to create similar pivotal moments across various fields, using this textbook case as a reference.

Consider this: had no one challenged Gillette up until now, they would surely be launching a ChatGPT integration for their razors, featuring enough blades to tame Bigfoot's mane, priced at $549.99. The "ChatGPT experience" will soon become commodified, and we will come to expect it as companies strive to provide it at minimal cost. Startups that merely utilize the ChatGPT API for specific use cases will struggle to maintain a competitive edge. Anyone can develop tools using the same API, and should they encounter obstacles, ChatGPT can assist in troubleshooting their code. It astonishes me that companies like Jasper.ai are valued over $1 billion, especially when they proudly claim their technology is simply ChatGPT applied to marketing tasks.

This leads me to contemplate whether generative AI represents a disruptive wave or merely a rising tide that lifts all boats. Ultimately, all paths lead back to the vast ocean of big tech. The choice lies in whether to build using the Microsoft-backed OpenAI API, Google's offerings, or the forthcoming Facebook versions.

Yet Google's attempts to engage with a lackluster chatbot remind me of retailers who anticipated the rise of e-commerce but viewed it merely as an extension of their physical stores. This outlook allowed companies like Amazon to step in with businesses crafted for the online shopping era.

All startups possess the fleeting advantage of starting from scratch with the finest technology available. Consider this: Would you create a search engine only to add a chatbot for optimal information retrieval? Would you invest months in producing static content for e-commerce sites, or would you opt for auto-generated content tailored to each user upon their arrival? Would you determine that existing project management tools are adequate, or could predictive AI enhance workflows and automate approvals? Would you rely on a sluggish ticket system for technical inquiries, or would an AI layer help categorize the queries and even resolve issues with minimal human input?

The list continues, and these questions are not the most profound we could explore. The answers are not immediately apparent, despite my framing them to guide you toward the latter options each time. There are inherent risks associated with generative AI that may justifiably prompt caution.

As technology evolves, such inquiries become essential starting points. On that note, Bill Gates recently joined the conversation, stating in a blog post that the emergence of generative AI represents the most "revolutionary" moment in technology since the advent of the graphical user interface in 1980.

Gates can afford to be mistaken, of course. He elaborates on the potential influence of AI on global well-being, particularly in health and education. In some regions, we might argue that the launch of the internet, smartphones, or even social networks has been more revolutionary than AI.

However, Gates emphasizes that while these technologies have been beneficial, their advantages are unevenly distributed. He notes that health and education standards — even in wealthier nations — are not improving at the same pace as technological advancements.

He states:

"In the next five to ten years, AI-driven software will finally fulfill the promise of transforming how people teach and learn. It will understand your interests and learning styles to deliver engaging content. It will assess your comprehension, recognize when your interest wanes, and identify what motivates you. It will provide instant feedback."

Gates’ five-to-ten-year timeframe appears cautious, highlighting the rapid progress we've made in a short span. More importantly, it showcases the advancements in technology. I can create a basic version of what Gates envisions using GPT-4, Python, and traditional HTML/CSS/JS today. In fact, I did so just yesterday.

We, the people, are still catching up to this technology, changing at a slower pace than the machines.

This oversimplified analogy often features in my investor pitch decks to illustrate the point: Generative AI relies on us. It requires our ideas, our content, our theories, and our experiences. While its creators may seek to shield its inner workings, we can be confident that AI's knowledge base is distinctly human. It is not some mystical or extraterrestrial force.

When we ask ChatGPT for ideas, we are not engaging with a higher being; we are merely interacting with a sophisticated program. If simply accessing existing content were all it took to enhance education, we would already be maximizing our current access to that content.

This is where Gates’ optimistic "five to ten years" projection makes several unwarranted assumptions. Achieving that vision necessitates collaborative effort with technology. We must bridge the growing skills gap that hinders our progress. Without this effort, while technology capabilities advance, our utilization of them may stagnate.

Every time I hear the phrase "the only limit is our imagination," I feel a shiver. Not because it is inaccurate, but because it is true. Our imagination may soon be confined to "ask ChatGPT for creative uses of ChatGPT," leading us in endless circles.

If we allow big tech companies to dictate the terms, they will constrict our possibilities to suit their existing frameworks. Google will continue to generate over $100 billion from its mediocre search experience while distracting us with superficial AI features.

However, if this truly is a revolutionary moment in technology, then big tech firms are akin to candle makers adding new fragrances while the potential of electricity looms around them. They are polishing the wheels of horse-drawn carriages while today’s innovators work on the automobile.

Until the next great breakthrough arrives.

Chapter 2: The Future of Generative AI

In this video, titled "Is ChatGPT-4o Actually Better Than GPT-4? OpenAI's Newest Flagship Model and Its Capabilities," we delve into the advancements and capabilities of OpenAI's latest model.

The second video, "OpenAI's NEW GPT 4o Just BROKE the Internet!" explores the significant impact of OpenAI's latest offering on the digital landscape.