Online: Trending Now

Too many of us are making snap judgments about AI Chatbots based on reading one or two reviews, or by taking only five minutes to test out the current version.

Truth be told, vendors are rushing chatbots out to the market, many of them in beta version, in order to meet their competition and secure a place for a much better version that is still under development. As we watch the growing number and variety of apps using generative AI, we are seeing an increasing breadth and depth of products. Yet, I daresay that none of them are near the potential that they will reach in a couple of years of refinement. So, when we discuss hallucinations and out-of-date databases, we should be careful about reaching summative judgments. These products are still very much in development; there will be new innovations and there will be bigger and better pools of data that will stir the pot among ranking brands and products.

As Steven Levy writes in Wired:

Yet it’s folly to draw definitive conclusions based on these early versions of the technology, including the shotgun blast of AI updates announced by Google this week. Folks, this is an illusion. Today’s chatbots are taking baby steps in a journey that will rise to Olympic-level strides. Oren Etzioni, former CEO of nonprofit research lab the Allen Institute for AI, told me this week that they’re already getting better. One force driving that trend is the millions of users—including reporters trying to goad the systems into doing nutty things that make good copy—exposing areas in need of improvement. Computer scientists, whose specialty is optimizing, after all, have been hard at work addressing the flaws.

So, what are we to do, and when will we know which generative AI apps are the best? Only time will tell which are the best for our particular needs. But, in the meantime, we should begin building our prompt engineering skills. I try to do at least half a dozen queries in a day spread among several apps. In most cases, I try to prompt at least two different apps with the same or similar prompt just to make the comparison and get a sense of what I can expect from each app. Yes, even these early versions are very different from one another.

The ones that I am testing for my own use include some of the popular ones and some which you may not yet have discovered. My choices tend toward the “free” apps with the exception of the “Plus” level that was required in the beginning to access language model GPT-4 using ChatGPT. Know that I am not invested in any of the companies in this report. Here are half a dozen of those apps that I regularly use.

ChatGPT, of course, is the early leader offered by OpenAI. I subscribed to the PLUS version for $20 a month to get preferential access to the latest version.

Bing AI Chatbot is associated with ChatGPT. It uses OpenAI’s most advanced Language Model GPT-4. You can automatically access it through a Microsoft logon. This chatbot has full, up-to-date internet access and it can generate images as well as text.

Perplexity remains one of my most-used apps. It is very fast, up-to-date, automatically generates citations of sources, and suggests follow-up prompts.

Pi, which stands for personal intelligence, is created by Inflection AI, a small AI start-up based in Palo Alto, California. Pi is a personal AI, designed to be useful, friendly and fun. It has a strong ethical framework and is committed to providing accurate and comprehensive information. This app will speak to you, using any of the four expressive voices from which you may select. It learns your unique interests and needs and provides you with information that’s relevant to you. It seems more concerned about you, the user, than other chatbots. There is the suggestion that it may promote good mental health.

Google Bard is a rapidly-improving app. After a faltering start, recent updates make it one of the best chatbots out there. It is up-to-date. Three versions of responses to prompts are offered giving different views or approaches to the prompt. The responses are well integrated into the other google online apps. I like the tone and detail I get in response to the prompts I write.

ChatPDF is an app of a different sort. It opens a pdf that you supply and allows you to use the power of generative AI to assess, analyze and improve the document. This is particularly useful in evaluating reports, assignments and research paper submissions.

Each of these has different features, and each is continuing to develop. They will improve over time. While we often speak of such apps using the term “ChatGPT,” generically, there are actually many varieties of generative AI apps with more to come. Sabrina Ortiz of Wired writes:

ChatGPT is only one of the many increasingly popular chatbots. Our picks for the best AI chatbots and writers can lighten your workload by writing emails and essays. Although ChatGPT has made quite the buzz, its popularity has made it unreliable for everyday use since it’s often at capacity. The good thing is there are plenty of AI chatbots that are just as capable, and available whenever you need them. We put together a list of the best AI chatbots and AI writers on the market and detailed everything you need to know before choosing your next writing assistant.

We will see improvements and additions to the list of available generative AI apps on a daily basis in the coming months. Meanwhile, we should practice writing effective prompts for the different apps you are following. Best practices will vary from chatbot to chatbot, yet the beginning guidelines should include giving precise background and context; stating expressly what you are seeking; and unlike prior search apps, more in the prompt is better than less. Write your prompts as you would speak to a professional colleague. The key to success is to practice, practice, practice until you can be confident in your prompting approach for that particular app.

 

This article was originally published in Inside Higher Ed’s Transforming Teaching & Learning blog.