Exploring the History and Future of Chatbots with Chat AI
IC the Future
Intelligent Conversational Agents Can Redefine Customer Engagement and Drive Success for Your Business
ChatGPT took the world by storm this year. Not only can it write text that sounds natural, it can answer many questions with a high degree of accuracy, remember what you say, execute commands and even write code.
The chat AI, introduced by OpenAI in November 2022, garnered the attention of the world, sparking debate, excitement and trepidation around the impact of AI.
While ChatGPT was the first chat AI program that was able to understand and generate truly natural-sounding language, it will certainly not be the last.
Given massive impact that these chat AI platforms are predicted to have on the world, it is critical for businesses and marketers to take a close look at the past, present and future of these tools.
How the Chat Landscape Has Evolved
Early attempts at chat interfaces far predate the recent ventures in chat AI.
In the 1990s, chatbots were deployed in text-based massive multiplayer online games (MUDs), AOL gave birth to SmarterChild, and Alicebot won three Loebner Prizes – awards provided to computer programs that proved to be the most “human-like.”
More recent chat interfaces, such as Microsoft’s Clippy and Tay have both been underwhelming, albeit for vastly different reasons.
Clippy was a pre-AI chatbot made in the image of a paperclip, whose attempts at assistance were considered by many users to be more intrusive than helpful. Microsoft chose to drop the paperclip, yet there are rumors that Clippy is about to return (if he ever truly left).
If he does, we can only hope that he will be more intelligent than and less corrupt than Tay, Microsoft’s AI-powered Twitter bot that, within 24 hours, began spewing racist, sexist and otherwise toxic language that it had learned from other Twitter users.
During this same era, many of us have grown used to voice-activated assistants that also use AI, such as Siri, Google Now and Alexa. Despite their ability to understand and process natural language, they cannot “chat” with you because they cannot actually generate new and original content.
Yet where other AI chatbots have failed, OpenAI’s ChatGPT has succeeded – so much so that many are predicting it represents the beginning of a new era.
What is ChatGPT + Why Should We Care
ChatGPT very well could represent an evolutionary step in generative AI, because it can very accurately simulate natural, “human-like” conversations with users.
Ask it a question, and it will provide a coherent answer.
Give it a command, and it will execute your order.
ChatGPT also has a memory, which means it can hold semi-coherent conversations.
These new capabilities are exciting enough, but the real reason ChatGPT is such a big deal is because it opens up entirely use cases and possibilities in business.
For instance, by itself, text-based chat AI can be used to automate:
- Customer service
- Technical support
- Sales
- Internet search
This last point is critical, as we have discovered this year: a chat-based interface would be much more efficient and user-friendly than Google’s current search box interface. The release of ChatGPT, in fact, reportedly prompted Google to declare “code red.”
But since chat interfaces are a poor vehicle for serving digital ads, which are Google’s primary source of revenue, it stands to reason that Google would not make the upgrade on their own.
That being said, ChatGPT is based on technology developed by researchers at Google, and Google has already trained a language model that is over five times larger than GPT-3, the language model upon which ChatGPT was originally built.
The clear and present danger that ChatGPT presents to Google’s business model almost certainly explains why Google introduced Bard A.I. in February, why they announced new AI-powered products at Google I/O 2023 in May, such as “Maya” – a tool that can visualize shoes in 3D to a wallpaper creation tool for Google Pixel phones and why Microsoft confirmed its billion investment into OpenAI.
How AI Chat is Shifting the Focus for Consumers, Marketers + Businesses
Fundamentally, what do all these changes mean for consumers, marketers and businesses?
If chat-based interfaces replace traditional search engines, we can expect a paradigm shift in the way people interact with the web and with brands. Rather than relying on a search engine to deliver web results or results from a technical support knowledge base, a chat AI interface could provide answers immediately.
Conversational AI is already being deployed on websites, in order to both fuel the sales process and acquire customer data.
Botco.ai is one example of a company that is using conversational automation for healthcare, senior living and other industries. By integrating Botco.ai’s platform within enterprise systems, organizations can immediately begin simplifying customer interactions and collecting data from chats, then use that information to inform other business decisions and activities, from marketing to customer experience management.
Other marketing software companies, such as Copysmith, Jasper, and Simplified have also begun to build upon OpenAI’s models to offer services that include content generation and image generation.
As AI evolves, we can expect to see even more advances, such as chat AI which auto-generates text, speech, and video. When conversational automation converges with other upcoming technology trends, such as AR and VR, it seems safe to say that the disruption we are witnessing today is just the beginning.
What is Yet to Come: The Future of Chat AI
Many of these changes are so abrupt, new and groundbreaking that many of us don’t know what to expect next. Will it enhance our productivity and economic output? Or will it put us all out of work?
When we look at the convergence of other impending technology trends, such as blockchain, the metaverse, and robotics, questions such as these become even harder to answer.
Perhaps the best people to answer these questions are the very experts who are building these solutions and investing in the AI-powered future.
- Chamath Palihapitiya, CEO and founder of the venture capital firm Social Capital, predicts that AI models such as GPT-3 will be commoditized and offered as a service, perhaps in the same way that GPT-3 is offered through a pay-per-use model today. While he says that ChatGPT and GPT-3 are pointing us in the right direction, these AI models will only become truly valuable when they can answer precise questions with overwhelming accuracy. Because they will be commoditized, proprietary data will become the key differentiator for AI-driven businesses: “Everybody’s chat model will eventually look and sound and feel like the same thing, unless you’re giving it a few special ingredients that other people are not.”
- Cathie Wood, CEO and CIO of Ark Invest, feels the same way. When asked about the hype surrounding ChatGPT, she reiterated that the companies that will benefit the most from AI will be those with proprietary data sets. According to her firm’s analysis, the productivity gains we see from AI will be “astounding and shocking,” but those gains will only go to companies that adopt and adapt fast enough.
- Sam Altman, CEO of OpenAI, also shares many of these views. Back in September of 2022, he predicted that chat AI would challenge Google’s current business model and that in the future, language models will go much further than people think. He also believes that multimodal models are coming, that models will be able to continuously learn (as opposed to GPT-3, which is “stuck” in the time that it was trained), that we will interface to our computers with chat AI, and that AI will ultimately become a foundational platform that will affect every industry. When asked about AI’s impact on the creative fields, he said that today’s models are only being used to enhance creativity, not replace it – but that in 100 years, AI may be able to do the whole job.
According to these sources, the near-term roadmap seems fairly certain: the advent of AI chat interfaces will have a massive impact on the way we work, interact with brands and do business.
At Ideas Collide, we remain cautiously optimistic about these trends and, for our part, we are already investing in the AI-powePreview in new tabred future and in chat AI with our partner, Botco.ai. Looking for more sight or to get started – we are here to connect!
“Everybody’s chat model will eventually look and sound and feel like the same thing, unless you’re giving it a few special ingredients that other people are not.”
-Chamath Palihapitiya