My 2023 ChatGPT SEO Predictions Revisited: What Came True, What Didn't

In early 2023, I wrote 9 predictions about how ChatGPT would destroy SEO as we know it. I compared it to the flip phone-to-smartphone moment. I said SERPs would become irrelevant. I said SEO was dead.

Two years later, it is time to hold myself accountable. Some predictions were spot on. Some were early. Some were flat wrong. Here is my honest scorecard.

Quick Context: What I Predicted in 2023

For those who did not read the original, here is the summary. I predicted that ChatGPT and generative AI would fundamentally change how people search, make traditional SEO practices obsolete, and force brands to rethink their entire digital strategy. I gave 9 specific predictions. Let me grade each one.

Prediction #1: SEO Is Dead

Original take: Traditional SEO -- content creation for E-A-T, backlink building, technical SEO -- all dead. ChatGPT would make it all irrelevant.

2025 reality: SEO is not dead. But it did change. Significantly.

Grade: C+. I was directionally right but too extreme. Here is what actually happened:

Content creation is not dead, but it changed. Google rolled out the "Experience" update (E-E-A-T, adding the extra E for Experience). AI-generated content flooded the internet. Google responded by rewarding firsthand experience and penalizing thin AI content. The bar for content quality went up, not down.
Backlinks still matter. They are less dominant than 5 years ago, but they are still a ranking signal. What changed is that earning backlinks through AI-generated content is harder because everyone is producing similar content.
Technical SEO is alive and well. Core Web Vitals, crawl optimization, structured data -- all still critical. If anything, technical SEO became more important as content quality equalized across competitors.

The lesson: SEO did not die. It evolved. The low-effort SEO tactics died. The practitioners who adapted to AI are thriving.

Prediction #2: How People Search Is Changed Forever

Original take: The old search loop -- type query, scan SERPs, click result, bounce back, repeat -- would cease to exist. AI would just give you the answer.

2025 reality: This one is playing out exactly as predicted, just slower than I expected.

Grade: A-. Here is the evidence:

Google AI Overviews launched in May 2024. For many informational queries, Google now generates an AI summary at the top of the page. Users get their answer without clicking a single result.
ChatGPT with web search became a legitimate search alternative. OpenAI reported 400M+ weekly active users by early 2025.
Perplexity carved out a real niche as an AI-first search engine with citations.
BUT -- most people still use traditional Google search for navigation, shopping, and local queries. The old search loop is not dead. It is just one of multiple ways people find information now.

The shift is real. It is just not as absolute as I predicted.

Prediction #3: SERPs Are Completely Irrelevant

Original take: Search engine result pages would be entirely removed from the UI. Ranking #1 on page 1 would become the new punchline.

2025 reality: SERPs are still very much alive.

Grade: D. I was wrong here. Google still shows organic results. People still click them. What happened is that SERPs got pushed further down the page by AI Overviews, ads, and featured snippets. Real estate above the fold shrank for organic results, but SERPs did not disappear.

The data: Organic CTR on position 1 dropped from roughly 27% to 19% between 2023 and 2025 for informational queries (per multiple industry studies). That is significant erosion, but it is not irrelevance.

Prediction #4: 50%+ Decline in Traffic to Websites

Original take: If AI handles informational and "everything else" queries (66.6% of all searches), website traffic would collapse by a similar percentage.

2025 reality: Traffic declined for some categories but not 50%+.

Grade: C. The numbers so far:

Informational content sites saw 15-30% traffic declines on average. Recipe sites, how-to sites, and simple Q&A sites got hit hardest.
E-commerce traffic stayed relatively stable. People still want to browse products, read reviews, and compare prices on actual websites.
Brand traffic increased for companies with strong direct audiences. If people know your brand, they still type your URL.
The big surprise: Social media referral traffic declined faster than search traffic for many publishers. TikTok and Instagram algorithm changes hit harder than AI Overviews.

I overestimated how quickly AI search would cannibalize web traffic. The decline is happening, but it is a slow bleed, not a cliff.

Prediction #5: AEO Is the New SEO

Original take: Answer Engine Optimization would replace SEO. The game would shift to getting your brand cited in AI-generated answers.

2025 reality: This is happening. AEO is real.

Grade: A. This might be my most accurate prediction:

GEO (Generative Engine Optimization) became an actual discipline. Companies now optimize specifically for AI Overviews, ChatGPT web search, and Perplexity citations.
New tools emerged. Platforms that track how often your brand gets mentioned in AI-generated responses, which sources AI models cite, and how to improve your visibility in AI search.
Brand mentions became a ranking signal. Not in the traditional backlink sense, but in the AI training data sense. If your brand is frequently mentioned in authoritative contexts, AI models are more likely to cite you.
The SEO tools adapted. SEMrush, Ahrefs, and others added AI search tracking features. They did not become irrelevant -- they expanded their product scope.

AEO is not a replacement for SEO. It is an addition to SEO. The practitioners who do both are the ones winning.

Prediction #6: The Value of SEO Is Trending Towards Zero

Original take: SEO's value would approach zero because AI would create abundant content covering everything.

2025 reality: The opposite happened for quality SEO.

Grade: D. Here is what I got wrong:

AI did flood the internet with content. Millions of AI-generated articles went live.
Google's response was to raise the quality bar. The March 2024 core update explicitly targeted low-quality AI content.
The result: Good SEO became more valuable, not less. Companies that invested in original research, expert content, and genuine E-E-A-T signals saw their rankings improve as AI slop got penalized.
SEO agencies that adapted to the AI era are charging more than ever because the skill required to rank went up.

The commoditized, low-effort SEO went to zero. The high-quality, expertise-driven SEO got more valuable.

For a practical look at what separates AI content that ranks from AI content that flops, see my AI SEO case study.

Prediction #7: ChatGPT Will Force Brands to Build Communities

Original take: Brands would have to shift from search traffic to community, social, and email marketing.

2025 reality: This is playing out.

Grade: B+. The evidence:

Newsletter growth exploded. Substack, Beehiiv, and ConvertKit all reported record growth. Brands that built email lists before the AI shift have a massive advantage.
Community platforms grew. Discord, Circle, Skool, and private Slack communities became key distribution channels.
Direct audience ownership became the top priority for smart marketers. Owned channels that do not depend on Google's algorithm or AI models.
However -- most brands still rely primarily on search. The shift is happening among early adopters, not the mainstream yet.

Prediction #8: Media Company Bankruptcies and Consolidation

Original take: Traffic decline would kill ad revenue, leading to media bankruptcies and consolidation.

2025 reality: This is happening, though AI is not the only cause.

Grade: B. What played out:

BuzzFeed shut down BuzzFeed News in 2023. Vice filed for bankruptcy. Sports Illustrated laid off nearly its entire staff.
Paywall adoption accelerated. The New York Times, The Atlantic, and others leaned harder into paid subscriptions.
AI content licensing deals became a new revenue stream. Major publishers signed deals with OpenAI, Google, and others to license their content for AI training.
The nuance I missed: The media decline was already in motion before AI. Social media algorithm changes, economic pressures, and declining ad rates were bigger factors in 2023-2024. AI accelerated the trend but did not cause it alone.

Prediction #9: Entertainment Content Is SEO's Last Stand

Original take: AI would master informational queries, leaving entertainment as the only reason people would visit websites.

2025 reality: Partially true, but the picture is more complex.

Grade: B-. What I got right:

Informational content did lose traffic to AI answers. "How to" and "what is" queries are increasingly handled by AI without a click.
Interactive tools, calculators, and personalized experiences -- things AI cannot replicate in a chat interface -- retained traffic well.
Video content (YouTube, TikTok) thrived because it provides an experience that text-based AI cannot.

What I got wrong:

AI is not close to mastering entertainment. It is not funny. It is not creative in the way humans find genuinely entertaining. That prediction about AI knowing "exactly what content to inject that dopamine hit" was too far out.
Product review content, comparison content, and transactional content still drives significant search traffic. These are not "entertainment" but they are not purely informational either.

The Predictions I Should Have Made

Looking back, here are the things I did not see coming:

AI-generated content spam would become a massive problem. The internet got flooded with low-quality AI articles. Google had to launch multiple updates specifically to combat it. Learn how to avoid spam detection on AI content.
AI image and video generation would progress this fast. DALL-E 3, Midjourney v6, Sora, Runway -- visual AI changed content creation as much as text AI. For more on Sora, see Sora explained.
The regulatory response. The EU AI Act, copyright lawsuits from the New York Times and others, and the ongoing debate about AI training data. This is reshaping what AI companies can and cannot do.
AI agents would emerge as a new paradigm. Not just chatbots answering questions, but AI agents that can browse the web, make purchases, and take actions autonomously. This has bigger implications for SEO than I originally imagined.

What Comes Next: 2025-2027

Based on what I have seen so far, here is what I think happens next:

AI search becomes the default for 30-40% of queries within 2 years. Not 66%, but meaningful enough to permanently reshape traffic patterns.
Zero-click searches continue to rise. Google will give away more answers directly. Website traffic for informational content continues its slow decline.
Brand becomes the most important SEO signal. If AI models cite you, if users search for your brand directly, if you have a recognizable name in your space -- that is what drives sustainable traffic.
The "Google-proof" business becomes a real strategic concept. Companies that built their entire acquisition strategy on Google organic traffic are vulnerable. Diversified traffic sources win.
AEO and GEO mature into standard practice. Every serious SEO strategy will include an AI search optimization component within 18 months.

In Conclusion

My overall accuracy rate: about 60%. Not bad for predictions made 48 hours after GPT-4 launched. The general direction was right -- AI is fundamentally changing search and SEO. But I overestimated the speed and underestimated how the industry would adapt.

The biggest lesson: technology shifts are never as fast or as absolute as the hype suggests. SEO did not die. It adapted. The practitioners who treated AI as a tool rather than a replacement came out ahead. The ones who panicked or ignored it got left behind.

The next 2-3 years will bring more change than the last 2-3. The playbook that works today will need updating by 2027. Stay adaptable.