In partnership with

Need a personal assistant? We do too, that’s why we use AI.

Ready to embrace a new era of task delegation?

HubSpot’s highly anticipated AI Task Delegation Playbook is your key to supercharging your productivity and saving precious time.

Learn how to integrate AI into your own processes, allowing you to optimize your time and resources, while maximizing your output with ease.

Get the free guide here.

OpenAI Wants You To Know Its New Model Will Be Underwhelming

OpenAI would like you to know that its new GPT-4.5 model, released yesterday, is going to suck.

That might seem like a strange marketing message from one of the world’s most exciting companies, but that’s essentially what OpenAI CEO Sam Altman said about the release of the 4.5 model.

On social media, Altman was quick to point out that the model is not a new frontier model. At best, it’s a small tweak—an iterative improvement to OpenAI’s flagship GPT-4 model.

“A heads up: this isn’t a reasoning model and won’t crush benchmarks,” Altman wrote when announcing the release. Altman also said “bad news: it is a giant, expensive model”

Why the weird hedging? It probably has to do with the fact that, as OpenAI has developed its most recent model, the sands of the AI landscape have been shifting under its feet.

Two years ago, when OpenAI started training its newest model, the name of the game was to build something as gigantic as possible.

GPT-3.5 was trained using an estimated 175 billion parameters, whereas GPT-4 surpasses 100 trillion parameters.

Many in the industry expected that further progress would come from making new models even bigger and more expensive.

Recent changes in the industry have called all that into question. Since the release of GPT-4, OpenAI rolled out its o1 model.

o1 is a reasoning model. That means that instead of relying on a gigantic size to answer questions, the model essentially takes its time, reasoning through its response before sending it to the user.

Reasoning models often return better results, and they rely more on processing time at the point where a user enters a query than on sheer size during training.

o1 initially looked less capable than GPT-4, but today it far outperforms its non-reasoning cousin.

Likewise, companies like DeepSeek have come along and shown that it’s relatively easy to distill a gigantic, expensive model like GPT-4 into a far cheaper, smaller one.

Size, in other words, is far less important for the power of a model than it used to be.

The trouble is that training new models like GPT-4.5—and the GPT-5 model expected to come out this summer—takes a long time. OpenAI likely started the process years ago. Now that they’re finally done, they’re embarrassingly in the position of rolling out a brand-new model that will likely underperform their better-tuned, more capable o1.

That’s the most likely explanation for why Altman is hedging. OpenAI knows it still needs to roll out new, large models, but it also knows that building bigger and bigger models is a strategy that’s already looking a bit dated.

Although we shouldn’t expect huge changes from GPT-4.5, many people in the content creation space will be happy with its new capabilities. OpenAI made it clear that the model is tuned to do a better job of understanding human emotion and communicating in a natural way.

That will make it a better writer and a better conversational partner for tasks like ideation, business planning, and more.

Even if the model isn’t smarter or better at solving high-level math problems and other benchmarks, GPT-4.5 may still prove exciting for the content creation slice of the AI industry.

The model is rolling out to users across ChatGPT over the next few weeks. I’ll share more as soon as I have a chance to get my hands on it.

How to Get Your Company Into AI Search Engines

Many content creators have spent years or decades learning how to do traditional SEO and marketing.

Now, AI search engines like Perplexity, as well as chatbots like ChatGPT and Claude, are quickly stealing market share from Google.

That begs the question, “How do you get your company or website into AI search?”

A major study just revealed key answers to that very big, very lucrative question. Jared and I talked all about it in today’s episode of the Niche Pursuits News podcast.