Further Contemplations on Email’s Future with a Nod to its History

As you may know from my last blog posting, “I have been meaning to make a blog post for quite some time about the history of email,” a task I never seem to get around to “because I always found the topic a bit bland.”

Well, I kind of got around to it in that blog post because I asked the then-current paradigm of artificial intelligence (AI), ChatGPT, to write the history for me. The AI program’s first stab at the task marked a barely passable start as a “history,” as it was about as light on historical detail as one could possibly get. Consider:

• Email was launched as electronic mail in 1971 by Ray Tomlinson, who sent “QWERTYUIOP” as the first-ever email message;
• Our usage of email evolved; and
• someone developed the “dreaded” email signature.

All three points are valid, but they hardly add up to the complete history of email. ChatGPT’s efforts were far more focused on eliciting a laugh than nailing down history—50 or so words vaguely referencing email history out of about 370 words total. Perhaps my fault, given that I requested a “humorous blog” that would “make people laugh.” Nevertheless, if you tell your audience upfront that what you are writing is “hilarious” and firmly suggest that you will “laugh out loud,” you are clearly not confident about your comedic chops. Frankly, ChatGPT’s efforts were mildly amusing at best.

AI Capabilities Continue Their Rapid Advance

Since posting that blog a few short weeks ago, ChatGPT has quickly evolved through several iterations to become now the fourth version, GPT-4, which “surpasses ChatGPT in its advanced reasoning capabilities.” While there is little doubt that GPT-5 may soon drop, interest in the AI space has shifted to Google’s Bard, touted by the media as ChatGPT’s new rival.

Naturally, I wanted to see how much Version 4 may have improved and was also interested in checking out Bard’s capabilities. Unfortunately, Bard is not “currently supported in [my] country,” so that examination will have to wait. To give ChatGPT credit, its latest iteration crafted a bit more actual email history—including references to the development of attachments and emoticons—when I requested another email history blog from it. Still, it was another history puff piece, but likely my fault as I again asked GPT-4 to compose the blog with hilarity and laughs. And repeat, it provided readers with upfront notice promising that what they were about to read would be “hilarious” and “laugh-out-loud” funny.

Not!

Its efforts remained mildly amusing, but GPT-4 did seem to make more of an effort to be funny than it did in the work produced by Version 1 a month ago . . . so, I guess it’s learning.

And My Long—Awaited Admittedly Bland—Email History Blog

To further my experiment, I decided to omit the humour and see what sort of facts-only email history blog GPT-4 might produce. Here are the results:

“Email has become a ubiquitous and essential part of modern communication, allowing people to quickly and easily exchange messages across the globe. But have you ever wondered about the invention, history, and evolution of email? In this blog post, we will explore the fascinating story of email and how it has transformed the way we communicate.

The Invention of Email

Email was invented in 1971 by a computer engineer named Ray Tomlinson. At the time, Tomlinson was working on a government contract for ARPANET, the precursor to the internet. He was tasked with finding a way to send messages between computers on the network, and he came up with the idea of using the @ symbol to separate the user’s name from the name of their computer. This allowed messages to be routed to the correct destination.

The first email message Tomlinson sent was a test message to himself, which read “something like QWERTYUIOP.” He later described the experience as “not very profound.”

The Evolution of Email

At first, email was a relatively niche technology, used primarily by scientists and researchers. However, as the internet became more widely available in the 1990s, email quickly became a standard mode of communication for businesses and individuals alike.

One of the key developments that helped to popularize email was the invention of the World Wide Web in 1989. This made it easier for people to access email, as they could now do so through a graphical user interface rather than using command-line prompts.

Another important development was the rise of mobile devices, which made it possible to check and respond to email on the go. The first smartphone, IBM’s Simon, was introduced in 1993, but it was not until the release of the BlackBerry in 1999 that email on the go became a mainstream phenomenon.

Spam and Security Concerns

As email became more widely used, it also became a target for spammers and hackers. Spam, or unwanted and unsolicited email, has been a problem since the early days of email. In the early 2000s, spam filters were introduced to help users filter out unwanted messages.

Security concerns also emerged as email became more popular. Email is inherently insecure, as messages are sent in plaintext and can be intercepted by anyone with access to the network. To address this problem, email encryption technologies such as PGP and S/MIME were developed.

The Future of Email

Despite the rise of other communication technologies such as messaging apps and social media, email remains a vital part of modern communication. According to a 2021 survey by the Radicati Group, there are over 4 billion email users worldwide, and the number is expected to grow to over 4.6 billion by 2025.

As email continues to evolve, we can expect to see new features and innovations that will make it even more useful and convenient. For example, some email providers are experimenting with machine learning algorithms that can automatically sort and prioritize messages based on their content and importance.

Conclusion

In conclusion, email has come a long way since its invention in 1971. From a niche technology used by scientists and researchers to a ubiquitous mode of communication used by billions of people worldwide, email has transformed the way we communicate. Despite the challenges of spam and security, email remains an essential tool for businesses and individuals alike, and we can expect to see it continue to evolve and adapt to the changing needs of users in the years to come.”

Other than omitting attachments, emoticons, and a few other important details, not too shabby as a rather bland, bare-bones email history article, and I’m pleased to get that knocked off my to-do list even though GPT-4 did it for me.

What Other AI Thinks About ChatGPT’s Efforts

As part of my experimentation with this new technology, I plugged the above-referenced GPT-4-generated history articles into another AI-supported program, Grammarly, which checks writing for grammar and other composition quality benchmarks. Grammarly was not overly impressed, but it’s typically not impressed with most people’s writing, so GPT-4 is actually doing OK with its Shakespearean endeavours. In checking the humorous blog, Grammarly identified two grammatical mistakes and described the article’s clarity as “a bit unclear,” its engagement a “bit bland,” and the delivery “slightly off,” all of which warranted an overall score of 83. As I noted, this is not too bad, as about 75% of documents checked by Grammarly fail to top scores over 80.

And ChatGPT continues to learn, given that its comedy-free effort showed marked improvement with a score of 86, with Grammarly describing it as “very engaging” and having “just right” delivery. Clarity remained “a bit unclear,” and the program identified six potential grammatical mistakes, though I only agreed with two.

I should add that Grammarly also passed both versions for plagiarism, highlighting 2% and 4% within respective texts for possible plagiarism. This made me ponder on this: Based on ChatGPT’s capabilities and capacity for prodigious output, is it possible that it could generate so many different compositions for a specific subject that it would be impossible not to plagiarize text for that subject?

But What of Email’s Future?

ChatGPT is successfully demonstrating that AI works and is rapidly increasing its capabilities and capacity for learning. It is also evident that AI will play a significant role in the future of email. And by the future, we’re talking near-term.

Observant readers of Atmail’s blog will notice that I posted an article about email’s future in December. That article—“What the future of email looks like (and why we need to invest in it today)”—short-shrifted AI’s game-changing potential by only referencing it with a couple of sentences. While blinders undoubtedly played a role in that oversight, the warp-speed pace of AI advancements has only become noticeably apparent in the past few months.

Chat developer, OpenAI, and Google are just two of thousands of companies directly working to enhance AI capabilities and bring AI-powered tools to market. While ChatGPT is not open-source software, OpenAI provides developers with API access to use and integrate into their own applications, which they are . . . in droves. And like OpenAI and Google, many companies—such as Microsoft, Adobe, and companies that didn’t even exist a year ago—are putting AI to use in the email realm.

With a quick Google search, you can find dozens of AI-powered email writers. “Stop wasting your time writing emails when AI can do that for you in seconds for free!” is just one of many similar tag lines. Or how about “Sky rocket your cold emails with unique personalizations generated by AI?”—Kind of like having the mind of C-3PO of “Star Wars” fame serving as your personal secretary, eh?

Along with email generation, companies are finding ancillary uses for AI in generating and managing emails. Shortwave, for example, uses AI to organize inboxes and summarize individual email content into a few short sentences, so you don’t have to bother reading the whole thing. While the tool seems to be getting rave reviews from the media—like this one—the unanswered question for anyone using it is what happens if the summary inadvertently omits a crucial component of a time-sensitive email? You know, something like, “and make sure the signed contract is delivered by 5 p.m. or the $10 million deal is negated.”

Or consider Mailbutler, which offers ChatGPT-powered email summaries and:

• Email auto-responses
• Email composition via short keyword input
• To-do list creation based on email content
• Contact info search and management
• Email editing and improvement

Other companies provide AI tools designed to help people quickly find images, videos, links, and other relevant input to add to their email messaging. And all these already-in-use ancillary email tools mark just the beginning of email’s AI-powered future. It almost seems like we’ll get to a point where we won’t ever have to open our inbox or compose an email again—just let C-3PO handle it all!

Not So Fast on the Demise of Email….

Operational AI may be on the market and has begun transforming how we manage our written words and email, but there are still “ghosts in the machine.” ChatGPT and similar smart chatbots periodically generate inaccurate or irrelevant responses and sometimes even “hallucinate” information lacking in real-world reality. Thus, reviewing AI-generated content for accuracy remains essential. “But Shortwave told me the signed contract wasn’t due until 7 p.m.!”

Even if you eventually end up with a personal AI-powered C-3PO to manage all aspects of your email usage, you’re still going to need email. As noted in my December “future of email” article, email remains the world’s preeminent communications platform despite the abundance of multiple other communications portals with similar features. It is also emerging as the most effective form of online identification, which is becoming essential for conducting business in today’s digital world.

Transformative technologies such as AI often—but not always—prove disruptive because of how they upend existing technologies and ways of doing things. Steam power certainly proved disruptive to most elements of industry, but the internal combustion engine that followed proved less so. The widespread adoption of email portended the potential demise of snail mail and government-operated postal systems. Mail volumes may be somewhat down, yet snail-mail delivery continues to operate globally.

AI might be on the cusp of transforming how we manage and engage with email, but it won’t disrupt our use and need for email. Over 100 billion email messages are sent and received every day. AI might represent a change in how the messaging is produced and managed, but it won’t slow the flow. Even if managed by a C-3PO-type secretary, just about everyone needs to maintain an email presence in today’s digital world.