GPT-5 Just Made SEO Irreplaceable

GPT-5 just made SEO irreplaceable. SEO is back. This is a tweet from Aleyda Solis. So Aleyda's Amazing SEO and Dejan SEO said GPT-5 just made SEO irreplaceable. So a couple bullet points here. OpenAI just made an executive decision to focus on raw intelligence and leave the rest to search engines. Now, number 2, without grounding. So grounding is grounding it on Maybe pulling the data from Bing or pulling the data from Google. So without grounding, this model is virtually useless. It's designed to be the brain on top of tools and information it's provided with. Number 3, and 4 bullet points here. This means SEO has never been more relevant than now. And the last point here is OpenAI appears to be designing models that think and reason well, but don't attempt to internally store all world knowledge. Instead, leveraging retrieval-augmented methods such as tools, search plugins, large context, to pull in fresh information as needed. And this approach is motivated by efficiency, cost, and performance considerations, as evidenced by recent statements, research, and product releases. Because they're burning so much money on these tokens, they have to figure out efficient ways to do this. So that's why I said SEO is back.

And people want cheaper prices for, you know, the APIs over time, and they have reduced their pricing quite a bit. I know they've lost a lot of money, but I think they lose money on purpose. They could charge more for all their products, but they choose not to, to try to gobble up more market share.

Yeah, totally. And I actually have some slides. So you and I are speaking at HubSpot's event actually in a couple weeks now. I'm going to work on our slides tomorrow a little bit, but check this out. So I have this. Okay, you're going to like this. Let me share this. You guys are getting a sneak peek to my talk. Neil, you're getting a sneak peek as well. Do you see this?

I do now.

Okay, so if you're just listening to this, the slide basically shows that the cost of intelligence is dropping. So I actually shared this at a talk I did earlier this year, but a couple months ago, you would pay for, let's say, like an input API price for like DeepSeek, right, which is a Chinese model, you pay $0.55 over here. And for ChatGPT's O1 model, which is the best in class at the time, You were paying $15 for that, okay? And then you can see on the very right, output API price is $60 from '01. Now today, look at what it's dropped to. So DeepSeek, maybe still 55 cents for this one, but look, instead of $15 for '01, it's dropped to $2. Instead of $60 over here on the right side, it's dropped to $8.

Well, '01 is $15 here, right? It's $2 for '03.

This needs to be adjusted over here. This O1 preview, ignore the orange. We're actually taking that out because no one really talks about O1 preview, but—

But right here, no, no, go to the next slide. It's O3 that's cheaper, right?

You see in the cash, it's $2.16. Yeah, yeah, yeah, yeah, yeah. Yeah, you're right. I didn't see this O3 pop in, so I actually need to make some adjustments to this slide, but point is, it's getting cheaper. So the fact that it's O3 on the right side here, this is— O3 was one of their best models at the time, and by the way, now it's like, it's just ChatGPT-5, so.

Yeah, but '01, you put $15 over time. Did it stay $15 or did it decrease?

I'm pretty sure it decreased, so that's why I need to fix the slide. But main point is this has dropped by a lot.

Yeah. So the pricing is all going downwards like we're talking about.

Yeah, exactly. We're just saying this is very deflationary and we're, we all, I think we all know, like you and I both know that, okay, energy is going to be really important. GPUs are going to be really important and we need the price to continue to drop down because we're just paying for intelligence ultimately.

Yes.

All right.

Yeah, I agree with that. You know, it's— you're just paying for intelligence at the end of the day.