Getting the Most Out of ChatGPT’s Advanced O1-Preview Model

OpenAI has taken a huge, major leap in the realm of AI — one that I’d call scarier than everything I’ve ever seen, in fact.

Their two newest preview models: o1-preview and o1-mini are crazy scary.

• o1-preview is able to perform similarly or even better compared to PhD students in chemistry, biology and physics — having the capability to seemingly reason through tasks which require thinking & problem solving (evident by its accuracy in such PhD-level questions)

• o1-mini is able to accurately generate and debug code much further than any other previous model (like GPT-4o).

Right now, we’re seeing a glimpse of the most advanced models at the tip of out fingers: with the ability to reason and think deeper than ever.

Here are some areas which I think it will revolutionize and significant use cases for the o1 models to be utilized:

Coding and Automation (even for non-coders)

Now, more than ever, the gap has been bridged for people who don’t know how to code but want to bring their actualizations to life.

Personally, I have standard coding skills to be honest, and am only familiar with basic syntax and logic. But with the help of models like GPT-4o, I'm able to bring my ideas to life with only basic and limited coding knowledge by simply understanding what I want to do, and carefully prompting GPT based on that vision, and then debugging errors.

(With this process, I've been able to create various applications with only basic coding knowledge, such as a Receipt Automator)

However, the new o1-mini model seems to be more powerful than the previous models, according to OpenAI: ranking in the 89th percentile on competitive programming questions via Codeforce.

Snapshot of OpenAI’s O1-Mini Page: https://openai.com/index/openai-o1-mini-advancing-cost-efficient-reasoning/

With the o1-mini model, non-coders can find ways to automate repetitive tasks and simple applications to streamline their workflows; especially with more complex visions.

Things like building applications for personal automation just got a lot easier: if you can think of and express an idea that you want (and with hours of debugging) you could create your very own automation toolkit with minimal coding experience.

Business Management & Planning

Instead of simply generating output, GPT o1-preview is able to “think” and reason better, and answer the problem with more intentionality and approach. It can dissect the problem step by step — something which is very useful.

As of now, however, I do think that its output may be a hit-or-miss, and for now, I do prefer the quality of some of the GPT-4o responses.

Short Snippet of O1-Preview’s Response

But I do think that we see from its output, that its way of thinking is much more intentional and approached. It’s as if it’s actually taking the time to think and reason, rather than just generating a response.

After all, this is just the “o1-preview” model, and the main o1 model is yet to be released (not yet public). I do think that definitely, that with its ability to reason and think critically on how to approach a problem (plus detailed prompting), this will be the tool to create full-blown business documents like:

• Targeted Business Plans (with SWOT Analysis)

• Targeted Marketing Plans and Strategies

• Targeted Content Creation Plans (Linkedin, Instagram, Facebook)

Previous GPT models, even the latest GPT-4o, have the capability to produce such documents: but what they’re lacking in is the ability to reason — sometimes, they’ll just respond with general and somewhat surface-level responses that can mainly feel “templated.”

But I think the o1-models could go even further and deeper: these models could produce actual quality documents that are targeted and thought in-depth.

The potential impact of this could be astronomical. With this development and reasoning as early as now, the automation of creating detailed plans, frameworks and proposals could be sped up and maximized through this new model (and with carefully targeted prompts, we have the opportunity to save so much time in productivity and free up our mental space to think big picture).