Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
OpenAI’s revolutionary chatbot ChatGPT has been all over the news in recent months, triggering technology giants such as Google and Baidu to accelerate their AI roadmaps.
ChatGPT is built on OpenAI’s GPT language model and provides a variety of functions, such as engaging in conversations, answering questions, generating written text, debugging code, conducting sentiment analysis, translating languages and much more.
Looking at the technologies of this moment in time, nothing seems to be as pivotal to the future of humanity as generative AI. The idea of scaling the creation of intelligence through machines will touch on everything that happens around us, and the momentum in the generative AI space created by ChatGPT’s sudden ascent is inspiring.
How should enterprise business leaders react to this? We thought that, by looking under the hood of ChatGPT and disassembling the application to its individual capabilities, we could demystify the product and enable any sufficiently-innovative enterprise to identify the elements most appropriate for their strategic relevance. Thus was born this analysis and research.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
We analyzed the various functions that ChatGPT provided and created an industry landscape map of the companies that fulfill one or more of these functions. You can think of this as dissecting ChatGPT into its various anatomical parts and finding potential alternatives for each function with its own unique and targeted capabilities. The resulting text generative and conversational AI Landscape is shown below and consists of ten functional categories with a sampling of representative companies for each category.
Breaking down the text generative and conversational AI landscape
Generative AI is a term rising in popularity with ChatGPT. It refers to AI technology that can create original content such as text, image, video, audio and code. Our landscape is focused on the area of text generative AI because that’s the predominant function of ChatGPT.
As you can see, the language models are at the bottom of the landscape because they form the fundamental building blocks of natural language processing (NLP) used for all the other functions. The sampling of language models shown here includes OpenAI’s GPT, Google’s LaMDA and BigScience’s BLOOM.
To the left of the landscape, we have grouped the categories of text summarization, sentiment analysis and text translation into the overarching category of text analysis, which refers to the process of using AI to analyze unstructured text data for patterns, insights and intent.
Text summarization companies use AI to summarize written texts into excerpts of the most important points. Companies in this category include QuillBot, Upword and spaCy. Sentiment analysis companies use AI to determine the emotions, opinions and tones inherent in written texts. Companies in this category include MonkeyLearn, Repustate and Cohere. Text translation companies use AI to translate written texts from one language to another. Companies in this category include ModernMT, TextUnited and Phrase.
Human-like interaction; code, text and search capabilities
In the middle of the landscape, we have grouped the categories of virtual assistants, chatbot-building platforms, chatbot frameworks and NLP engines into the overarching category of conversational AI. This encompasses technologies that interact with people using human-like written and verbal communication.
Virtual assistant software responds to human language and helps the user with a variety of tasks and queries. Companies in this category include Augment, Replika and SoundHound. Chatbot-building platforms enable non-technical users to create and deploy chatbots without writing code.
Companies in this category include Amelia, Avaamo and Boost AI. Chatbot frameworks and NLP engines enable developers to create chatbots using code, and also build the core components of NLP. Companies in this category include Cognigy, Yellow AI and Kore AI.
To the right of the landscape, we have the categories of writers, coders and search. Writers use AI to create original written content and edit existing written content for grammar and clarity. Companies in this category include Jasper, Writesonic and Grammarly.
Coders use AI to generate code from natural language inputs and debug existing code. Companies in this category include Tabnine, Replit and Mutable AI.
Finally, search comprises AI-based search engines for the entire web or for an enterprise’s internal knowledge base. Companies in this category include Neeva, Perplexity AI and You.com.
The ten categories
- Text summarization: These companies use AI to identify the most important information from long form texts and summarize them into short digestible excerpts. Other functions of these companies include keyword extraction, text classification and named entity recognition.
- Sentiment analysis: These companies use AI to determine the sentiment of the text as either positive, negative or neural, as well as the tone, emotion and intent behind the text. Sentiment analysis is often used in analyzing customer feedback and brand attitudes.
- Text translation: These companies use AI to translate text from one language to another, mostly for written text but also for voice and video recordings.
- Virtual assistants: These companies create voice-enabled or text-enabled assistants that help the user with a variety of tasks such as taking notes, scheduling appointments, recommending products and providing mental health therapy.
- Chatbot building platforms: These companies provide an interface for non-technical users to build and deploy chatbots without needing to write code. They usually include a visual builder to designate the flow of interaction with the chatbot.
- Chatbot frameworks and NLP engines: These companies provide an environment for developers to build and deploy chatbots using code, as well as companies that build the core component of natural language processing which converts human language into machine inputs.
- Writers: These companies use AI to generate written text for given topics such as essays, poems, blog posts and sales copy. They also help edit and paraphrase written text for grammar, tone, clarity, and style.
- Coders: These companies use AI to assist developers in generating code from natural language descriptions. They also help debug existing code and explain the reasoning behind their code edits.
- Search: These companies use AI to search the web for answers to questions about general knowledge, as well as companies that build custom search solutions for an enterprise’s own internal knowledge base.
- Language models: These models learn from an abundance of human written and spoken texts, and predict the probability of the next word in a specific sequence of words. They form the fundamental building blocks of NLP used for text generative and conversational AI.
Broad landscape, evolving challenges
As you can see, the landscape of functions similar to ChatGPT is broad, with a growing number of companies competing in each function. This infographic shows only a fraction of the 700-plus companies we have uncovered in the space, with more products and companies launching daily. Similar to other major technology shifts we have seen with the internet, mobile, and more recently in crypto, this early spring tide of market buildup consists of an explosion of activity that will continue to accelerate before shaking out and consolidating in the years to come.
The obvious challenge for enterprise leaders in this phase of the market evolution will be navigating the landscape and identifying the true signals. What are the opportunities that can accelerate their businesses, provide new value to their customers or keep them competitive in a rapidly changing market?
Facing the plethora of competing generative AI products, enterprise leaders need precise criteria for weighing and selecting the right ones for their creative and knowledge workforce. It may turn out that a portfolio of solutions would work best, and the role of knowledge and creative workers evolves from creating original content to comparing, collating and editing the best creative output from the multitude of generative AI tools. One thing is for sure; every enterprise must have a generative AI plan.
Dong Liu and Nader Ghaffari are co-founders at Daybreak Insights.
Special thanks to Arte Merritt for his review and feedback.
Welcome to the VentureBeat community!
DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.
If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.
You might even consider contributing an article of your own!