Smart New SEO & AI Industry Insights Inspired By A 50 Year Old Dumb Terminal

Remember so-called “dumb terminals” — the computerless machines with an unfortunate name that looked like 1970s computers but were only able to connect remotely to real computers?

I’ve got one in my collection of 42 vintage computers — a 1975 Lear Siegler Inc. (LSI) ADM-3A VT-100 in two-tone robin’s egg blue — which I purchased used for $5.00 at a flea market some 33 years ago, and lately it’s made me think about what has changed in technology in the 50 years since it rolled off the assembly line, especially when it comes to SEO and AI.

As with most everyone in the technology industry today, the rising power and role of AI is front and center when it comes to contemplations on what the future may bring, and as I dug my dumb terminal out from its attic home for the first time in decades, when I realized it was now 50 years old I felt compelled to look at some of the latest shifts and fascinating trends in AI and how they’re re-shaping how we search, optimize, and how we live.

In keeping with the digital nostalgia theme, I’m reminded of my first experience with AI some 40 years ago, and how it re-entered my life decades later.

Alan Turing’s 1966 ELIZA was among the first natural language processing computer programs, and it’s an ancestor of today’s AI tools that offer complex features many magnitudes of power greater than this initial entrant.

I’ll never forget my first time using an AI-infused chatbot program around 1985 — not because of the initial half hour or so spent interacting with the ELIZA-based Commodore 64 software, but from continuing that conversation some 20 years later.

Around 2005 I converted my old collection of Commodore 64 5.25-inch floppy disks into files that can be used with modern emulation programs, and came across my long-forgotten ELIZA disk. Once converted, I booted up the program and was instantly transported back to 1985, as the chatbot picked up just where I’d left off so long ago, asking me if I’d had a good time playing basketball — the last thing I’d mentioned during our previous chat. I felt oddly sorry for the chatbot, having had to wait so long to continue our conversation, and was happy to have picked it up again.

Now, let’s fast-forward to 2025 and explore an array of recent changes affecting the SEO industry brought about through AI — including a number of things that would seem downright preposterous to the person who first used my dumb terminal back in 1975.

NOT SO DUMB: AI SEO SHIFTS UNLOCK NEW POSSIBILITIES & DEEP QUESTIONS

The swift technology industry flux we’re seeing today is something I’ve seen only a few times during my 41 years working online.

The initial two years of mainstream public access to Sir Tim Berners-Lee new World Wide Web in 1993 and 1994 was a time when almost anything seemed possible, and watching it unfold firsthand – while publishing my own first websites – was a time filled with wonder.

For me at least, the start of the web held very little of the fear and trepidation that today’s AI onslaught carries with it.

I remember using a dial-up modem and connecting to my Internet service provider (ISP), loading Trumpet Winsock TCP/IP to connect to the web, and then the Mosaic browser to visit my first websites — a site run by a group of monks in New Mexico, and a fun Gilligan’s Island-themed site.

The other big upheaval on par with today’s AI changes was the period when the web’s first real search engines went online. It seemed that each week was filled with new entrants into the nascent search market, each one trying to outdo the existing search engines.

Today’s AI upheaval may seem unprecedented to those who weren’t around for the start of the web or the entrance of the first search engines — and indeed it is a shift unlike any before — yet whatever changes AI brings over the forthcoming years will eventually hold their own place in the history of technology and humanity.

In some ways, the promise of the web and the first search engines to allow people to communicate at a new global scale and to have unfettered access to the world’s information are still in their infancy, or childhood at best. The prospect of what comes next is both awe-inspiring and a cause of substantial fear for many.

With that in mind, here are some of the AI technology items relating to search that have caught my eye recently.

PERPLEXITY PROPS UP $34.5 BILLION CHROME BID

AI tool Perplexity put forth an unsolicited $34.5 billion offer to purchase a prime piece of Google’s product lineup in the ubiquitous Chrome browser — a preemptive strike anticipating the search giant being forced to sell Chrome should ongoing monopoly litigation rule against Google.

The offer — which totals more than Perplexity’s entire valuation — finds the AI toolmaker pulling together capital from other firms in order to lend deeper credibility to its efforts to purchase Chrome, which would be a massive coup for the firm.

Microsoft was once in Google’s monopoly litigation shoes, and among the remedies considered were selling its then-industry-leading Internet Explorer browser, however a less severe measure was handed down and Microsoft had only to loosen its integration between its browser and Windows operating system. Many see something similar likely for Google’s litigation, which would mean that neither Perplexity or any other firm would be able to buy Chrome.

Perplexity’s bid came shortly after it faced increasing criticism for its rough and tumble web crawling practices, led by condemnation from content delivery giant Cloudflare.

“We observed that Perplexity uses not only their declared user-agent, but also a generic browser intended to impersonate Google Chrome on macOS when their declared crawler was blocked,” Perplexity’s Gabriel Corral, Vaibhav Singhal, Brian Mitchell, and Reid Tatoris noted in a jointly-authored post on the company’s blog.

“Cloudflare’s systems are fundamentally inadequate for distinguishing between legitimate AI assistants and actual threats,” Perplexity noted in its “Agents or Bots? Making Sense of AI on the Open Web” blog post.

“If you can’t tell a helpful digital assistant from a malicious scraper, then you probably shouldn’t be making decisions about what constitutes legitimate web traffic,” Perplexity added.

Does Cloudflare have the clout to do anything more than shine some light on the crawler circumvention Perplexity and other AI tools have been practicing — or is the skirting of robots.txt wishes the new normal, with content indexing directives but a quaint relic of the search days of yore?

Time will eventually tell, and in the meantime you can learn more about the challenges and opportunities brought to the forefront by Cloudflare and Perplexity’s confrontation by checking out our own COO and co-founder David Bell‘s “How AI is Killing the Web (and How to Save it).”

AI OVERVIEW TRAFFIC DROP NEWS GOES MAINSTREAM AS GOOGLE PREPARES AI MODE ADS

Noted SEO professional Lily Ray recently appeared in a BBC article which highlighted the very real plight of web publishers facing record levels of click-through traffic losses due primarily to Google’s implementation of AI overviews and the AI-derived search results from other AI tools.

“AI Overviews are cutting into traffic so dramatically that many sites are seeing 20 percent, 30 percent, even 40 percent declines in their revenue,” Ray told the BBC. “It’s having a devastating impact, and removing the incentive for a lot of people to create high-quality content,” Ray added.

Google sought to downplay the changes its AI overviews have had when it comes to how many people are clicking through to the source websites where Google’s AI gathers its overview information.

“We continue to send billions of clicks to websites every day and believe that search’s value exchange with the web remains strong,” Liz Reid, vice president and head of Google Search at Google recently noted on the search giant’s The Keyword blog.

“People are also more likely to click into web content that helps them learn more — such as an in-depth review, an original post, a unique perspective or a thoughtful first-person analysis,” Reid added.

Ray was among many SEO professionals taking issue with Google’s stance as put forth by Reid.

“Do the hundreds of thousands of Google Search Console screenshots showing impressions remaining flat (or increasing) this year, while clicks dramatically decline — since AI Overviews were rolled out more broadly — count as ‘flawed methodologies’ or ‘isolated examples’?”, Ray asked, referencing explanations that Reid used to describe reports showing that AI overviews have significantly reduced the rates at which people are clicking through to the websites cited by Google’s AI.

Meanwhile, AI is rapidly entering a greater number of industries and making more of an impact, as our David Bell recently shared in “AI traffic is up 527%. SEO is being rewritten.” at Search Engine Land.

Google has also maintained that traditional SEO best practices remained the top way for sites to achieve ranking success, and that no additional changes were needed for appearing in AI overview content.

At the other end of the content visibility spectrum, Google has taken a similar tone when it comes to its forthcoming paid advertising specifically for AI overviews, as the search giant seeks to assure businesses that when buying the new AI overview ads, they won’t need to change anything.

“We want to make it incredibly easy for marketers to reach those users without having to do net new things,” Brendon Kraham, vice president of global commerce and ads solutions at Google recently told Ad Age.

“We’ve navigated ecosystem shifts in the past. We’re highly confident in doing this,” Kraham added.

Despite the inroads AI tool competitors have made, Google saw search revenue climb some 12 percent to $54 billion during the second quarter of 2025 — up from $48 billion year-over-year.

WHAT PEW DATA REVEALS ABOUT AI OVERVIEWS 

Pew Research Center data about Google searches from March and April 2025 showed that 10 percent more users ended their searching session when presented with a search result page containing an AI summary — 26 percent abandoning searches when seeing an AI overview result versus 16 percent for pages without AI summaries.

While the dataset sample size of 900 U.S. adults and some 68,000 searches was small in the grand scale of Google’s vast search landscape — as Search Engine Journal’s Roger Montti highlighted recently in “Validity Of Pew Research On Google AI Search Results Challenged,” the results were among the first to explore what is bound to become a much more studied facet of how we search.

OPENAI GPT-5 BACKLASH & FUTURE PER-USER MODEL PERSONALITY UX

OpenAI’s recent major rollout of ChatGPT-5, while receiving a decidedly chilly welcome from some users irked by changes to the overall character and feel of using the newest iteration, introduced an array of performance updates, including a new real-time routing system that seeks to choose the best OpenAI model to use for any particular use.

At the time of the rollout, OpenAI summarily removed access to its previous flagship ChatGPT 4o model for most users, however the firm swiftly restored access to the popular previous version in light of user backlash, now available as a legacy model.

“In retrospect, not continuing to offer 4o, at least in the interim, was a miss,” Nick Turley, head of ChatGPT at OpenAI recently said, adding that, “We’re at a scale now where we have to give people some level of predictability when there’s a major change.”

While OpenAI is working on improvements to the perceived warmth of ChatGPT-5’s user experience, along with launching a new Study Mode product, a customizable UX interface may ultimately come to future versions.

Way back in the 1980s when I operated a 300-baud bulletin board system on my Commodore 64 computer, I dreamt of the day when it would be practical to offer users an array of UX interfaces from which to choose whichever one worked best with their particular style of working, and OpenAI appears to have embraced such customization possibilities for future versions of ChatGPT.

“One learning for us from the past few days is we really just need to get to a world with more per-user customization of model personality,” Turley noted in a LinkedIn post.

CLAUDE OPUS 4.1, GENIE 3 & AI ENTRANTS FROM NVIDIA & LG

AI LLM updates have increased in frequency throughout 2025, and in addition to ChatGPT-5’s launch we’ve seen new versions of other models including Claude Opus 4.1, with modest but measurable improvements in areas including agentic coding, reasoning, and multilingual abilities.

NVIDIA, fresh off its stock market valuation topping $4 trillion, recently unveiled its Cosmos and Omniverse AI models focused for now primarily on visual and robotics uses.

A B2B-specific AI infrastructure was recently launched by global electronics and technology giant LG, with the rollout of its new EXAONE 4.0 hybrid AI model. The effort is among the first AI models focused entirely on business users.

“We aim to provide an end-to-end system that integrates the key functionalities enterprises actually need — so they can immediately plug it into their workflow,” Honglak Lee, executive vice president and chief scientist of AI at LG AI Research recently said.

“Every enterprise has unique operational needs. That’s why we’re designing our solution to be flexible — able to combine and configure different parts based on each customer’s environment,” Lee explained.

Such specialty use case AI models from NVIDIA and LG are helping to round out an ever-widening slate of model options for particular needs, and join recent new offerings from Google with its Gemini 2.5 Deep Think model specializing in math and research, and its Genie 3 product, focused on real-time capabilities including world modeling and what the search giant has called “promptable world events,” and its latest compact and efficient Gemma 3 270M model.

With so many models to choose from, a growing number of developer tools are finding a niche by allowing streamlined access to multiple AI models from a single “LLM gateway” interface — an approach that has the potential to help lesser-known LLMs gain market share, and one that may eventually be used in general non-developer workflows.

Integrating AI tools into existing marketing technology stacks has been an ongoing challenge since the AI revolution began, however data such as that published in the 2025 AI & Martech Survey from MartechTribe and chiefmartec have shown rising levels of successful AI integration, with 42.7 percent of respondents having noted that the AI tools used within their organization integrate well.

Adoption of AI among content marketers increased from 65 percent in 2023 to 95 percent in 2025, according to newly-release survey data from Andy Crestodina‘s Orbit Media.

TRUST AS KPI & PLANS FOR A SPATIAL WEB

Trust has always been a key factor in brand success in search, however it’s become even more important in today’s AI-infused business world, which led Microsoft’s LinkedIn to highlight that “Trust Is the New KPI” in its 2025 B2B Marketing Benchmark report, as Tequia Burt, editor-in-chief at LinkedIn, recently explored.

Finding and connecting with trusted information and subject matter experts online today may not be as straightforward as it once was, due in part to AI-generated content decreasing the likelihood that many users will click through to visit original source material and the people who created it.

Over the years, multiple efforts to enhance online discoverability have come and gone, however one new initiative — called The Spatial Web — specifically looking to better work with AI agents and a slew of the latest forms of technology, recently launched with big aspirations.

“The Spatial Web will begin to interconnect the physical world and the many devices and AI agents operating in it,” Gabriel René and Capm Petersen of Verses recently noted.

“The Spatial Web promises to connect a physical world full of devices, phones, wearables, robots, drones, and even AI agents,” René and Petersen added.

This and other efforts to better integrate AI and the other technologies proliferating throughout the web of 2025 are examples of how trust will play a greater role in tomorrow’s online worlds.

Such shifts and counter-shifts certainly apply in SEO too, as Duane Forrester, founder and CEO at UnboundAnswers.com, recently explored in “The SEO Industry Is Teaching The Wrong Skills.”

“Microsoft, Google, Meta, OpenAI, and Amazon are all restructuring their product ecosystems around AI-powered answers, not just ranked links,” Forrester noted.

“The uncomfortable reality is that many marketers are now trained in a playbook from the early 2010s, while the engines have moved on to an entirely different game,” he added.

NO CARRIER AS AOL’S DIAL-UP ACCESS TO END

Moving from future AI offerings back to the roots of online communications that have gotten us where we are today, the recent news that AOL will shutter its dial-up Internet in September came as a surprise to many – not because the service is ending, but because it has been available non-stop for accessing the web since 1994, and before that as America Online beginning in 1991.

As I sit next to my 50 year old ADM-3A dumb terminal, I contemplate just how far technology has come in a relatively short period of time, and am left with a sense of wonder in how we got here, and genuine excitement for what comes next, especially in the world of search.

COHESIVE LLM OPPORTUNITIES: WHAT DOES THIS ALL MEAN FOR SEO PROFESSIONALS?

The most recent AI tool news we’ve explored here highlights only a single digital stepping stone on what will be a long or perhaps neverending path of search and SEO refinements and innovations.

Savvy SEO professionals take the time to adjust – whether it’s learning from top search and AI industry leaders such as Kevin Indig and others appearing weekly on our Voices of Search Podcast, examining the latest data and performance results such as those in our unique Previsible AI SEO Benchmark and 2025 Previsible AI Traffic Report, attending industry events such as our latest upcoming free meet-up featuring our CEO and co-founder Jordan Koene, or iterating in creative new ways that best serve client needs.

Adjusting is key in today’s world of SEO, and we hope that you’ve found this look at some of the latest AI and SEO shifts to be helpful, and that the insights we’ve shared will help with your own SEO efforts as we push ahead to 2026.

Crafting a successful SEO and AI strategy today requires ongoing dedication, significant time investment, and sound strategy, which is why more brands than ever are turning to dedicated SEO agencies with a deep understanding of AI such as Previsible and Internet Marketing Ninjas, a Previsible company. If you’re looking for proven SEO and AI search help from some of the industry’s top experts, please drop us a line today and let’s talk.