Generative AI chatbots have rapidly become indispensable tools across various industries, transforming the way we interact with technology. These advanced platforms are not just for chatting anymore; they’ve evolved into multi-modal systems capable of understanding both language and visual information. This makes them some of the most sophisticated AI tools available today. As the market continues to grow and evolve, new and innovative chatbots are being developed at an unprecedented rate, offering enhanced capabilities and functionalities.
In this article, I want to highlight five notable generative AI chatbots that stand out for their unique features and the broad range of tasks they can perform. From aiding in coding and writing to generating images and even engaging in complex conversations, these chatbots represent the forefront of AI technology, demonstrating the incredible potential of generative AI in various applications.
ChatGPT
ChatGPT wasn’t the first generative AI tool to make it to public release. Those of us with an interest in AI have been playing about with image generators and such for a few years already. But it was the first to really impress upon a wider audience just how ready AI was to hit the mainstream.
Within weeks of launching, it hit one million active users. It was widely reported that this was the fastest-growing audience for any app ever – although this record was broken shortly after when Meta launched Threads.
Over the year since it was originally released, OpenAI has worked hard to keep us interested. First, they launched a Pro version powered by its latest and most powerful large-language model (LLM) GPT-4. Then, it added web browsing capabilities and image generation powered by Dall-E, making it truly multi-modal.
ChatGPT is often referred to as the “do-anything-machine” as it’s a great first port-of-call when you want to get just about any job done. If it can’t do it for you itself, there’s a pretty good chance it can tell you how to do it yourself. ChatGPT is the original and, in many ways, still the best. Most people who’ve used all of the tools listed here will probably agree that as a general-purpose workhorse, ChatGPT is at the front of the field.
Google Bard
Many of us thought Google – the reigning champion in the world of AI – had been caught on the back foot by the arrival of ChatGPT.
When ChatGPT emerged, it was immediately recognized as perhaps the first serious threat to Google’s long-term dominance of the search industry – the source of the majority of its revenue.
The response was Bard, which took a while to arrive and at first looked like a pale imitation of OpenAI’s upstart chatbot. However, coming up to a year from its release, it’s evolved to become capable and useful.
Unlike OpenAI, Google has jumped between models behind the scenes. At first, Bard was powered by LaMDA before a newer model, PaLM 2, was introduced, improving its coding and mathematics abilities. The latest switch was to Gemini Pro, with a future upgrade to Gemini Ultra in the works. Google’s latest version is even reported to outperform GPT-4 in some tasks, such as speech recognition. The most recent updates have given it image generation capabilities powered by its Imagen 2 technology.
One advantage that Bard has over ChatGPT, for some at least, is how smoothly it slots into the Google ecosystem. If you’re a user of Gmail, Workspace, Documents, and so on, you’ll be impressed by how easily it’s able to draft emails, create documents, generate data, or automate many routine tasks.
Microsoft CoPilot
Copilot is Microsoft’s current name for its flagship AI chatbot, which launched as a new version of its Bing search engine called Bing Chat, before acquiring its own name and independent identity.
Microsoft have chosen the name carefully, to convey the feeling that it’s intended to help us, rather than simply chat to us. By integrating AI across all of its work and productivity tools like Windows and Microsoft 365, it hopes to become the mainstream choice in AI, just as it has done in those markets.
As a major investor in OpenAI, Microsoft has privilege when it comes to using its technology in its own products. The original Bing Chat was the first opportunity many of us had to experience GPT-4, and the most powerful all-round LLM is the backbone of CoPilot today. Like ChatGPT, it also uses Dall-E to generate images.
But its real advantage is that it injects AI into tools that millions of us use every day. Spreadsheets, text documents and computer code can be created with natural language prompts. It’s widely used by coders due to its integration with the Github coding platform, also owned by Microsoft.
Llama2
Meta’s answer to ChatGPT is its multi-modal Llama2 model. However, rather than package it as a commercial product, like Microsoft or OpenAI, it’s taken a slightly different approach. Following a quasi-open-source licensing model, the code and training data are available for anyone to use to create their own chatbots. It can also be accessed through its own URL if you aren’t a developer and just want to find out what it can do.
Meta has said that it’s taken this approach to make Llama as accessible as possible. One advantage is that it enables private instances to be created that don’t have to send data back to Meta or the cloud for the AI to access it. Because of this, although it can be considered a general-purpose AI chatbot, in the same manner as ChatGPT or Bard, it is seen as particularly useful for building more specialized applications. There are several open-source LLMs available now, but (according to its own tests) Llama2 outperforms them all.
Claude
Claude has been created by Anthropic, a company that was started by former OpenAI employees. It is the first multi-modal chatbot they’ve built, capable of handling text, voice, images, and documents. Users say that they find it fast and capable and that it generates highly coherent responses. However, it’s somewhat narrower in scope than ChatGPT or Bard when it comes to what it can do.
Anthropic has stated its commitment to ethical and transparent AI, which is reflected in a principle called Constitutional AI. This has resulted in a chatbot that’s uniquely capable when it comes to engaging with users who (perhaps unknowingly) ask it to generate content that could be unethical or harmful. It can explain the rules it follows, give reasons for its behavior, and suggest alternative ways to accomplish tasks without crossing its guardrails.
Since it launched hot on the tails of ChatGPT in early 2023, Claude has stood out due to the fluency of the conversations it can hold and its ability to understand subtle nuances and differences in the ways that humans communicate. It also lets users assign it a persona that they like and find pleasant to talk with.
Read the full article here