The AI Era: 7 Cognitive Upgrades and Optimizations for B2B Product Managers
-
Recently, the news of former OpenAI CEO Sam Altman being fired by the board has been making waves.
It even overshadowed Altman's newly announced project: GPTs, which stands for custom GPTs.
GPTs further amplify its value of 'understanding human language':
In the past, you could generate an article just by speaking.
Now, you can generate an AI product just by speaking.
As GPT becomes increasingly powerful, could it potentially replace B-side product managers?
For example, users might directly utilize GPT (rather than going through a product manager) to generate B-side applications.
If this is the case, where does the value of B-side product managers lie in the GPT era?
How can B-side product managers hop on the GPT bandwagon?
Here are 7 insights I’ve gathered after discussing with AI SaaS practitioners, which I hope will inspire you:
GPT's strengths lie in its understanding of human language and its ability to generate intelligent content.
However, its shortcomings are equally evident: reliance on online data, opaque logic (black-box nature), and the phenomenon of content "hallucination."
In contrast, the core responsibilities of B2B product managers revolve around offline business research, rigorous logical analysis, and, based on these, accurate requirement understanding and product delivery.
These are areas where GPT falls short.
Therefore, GPT will not replace B2B product managers; instead, it will highlight their unique value.
Although GPT is more intelligent and even has the potential to revolutionize the way humans interact with the internet.
However, like apps, GPT is still just a tool, and the barrier to using it will continue to lower.
In comparison, mastering business scenarios and understanding customers and industries is much more difficult than mastering GPT.
Just as in the mobile internet era, familiarity with 'app interaction' did not become a core competency for B2B product managers.
Therefore, the emergence of GPT may lead to an "upgrade" in B2B products, but it won't cause a "replacement" of B2B product managers.
It can even be confidently stated that in the future, B2B product managers with a profound understanding of business and operations will become even more sought after.
Before the widespread adoption of mobile internet, there were actually very few B2B product manager positions in China.
After the popularization of mobile internet, as many offline business scenarios had the opportunity to go "online," it led to an explosion in apps and a surge in demand for B2B product managers.
With GPT's capabilities continuously improving, costs decreasing, and new use cases being developed, it will inevitably lead to another explosion of internet products.
Therefore, although the future may not necessarily require more programmers, there will certainly be a greater need for B-side product managers.
Having some technical knowledge is certainly better, as we need to communicate with technical colleagues, and a lack of common language can affect communication efficiency.
Especially for traditional AI product managers, if you don't understand algorithms, it's difficult to design AI products that meet requirements.
However, GPT's advantage lies precisely in lowering the technical threshold.
For example, its ability to understand human language enables us to generate applications with complex algorithms simply by "speaking" (rather than writing code).
This means that as long as we understand GPT's capabilities and limitations, as well as the scenarios it is suited for, even without knowledge of algorithms, we can leverage GPT to build AI products that meet our needs.
To confirm this, I specifically sought the opinion of Mr. Yan, co-founder of AI SaaS startup Quick Creator.
He believes that designing GPT-based AI application products primarily requires understanding the business domain, but not necessarily algorithms or even a technical background.
The crucial first step is to become a heavy user of GPT.
Just as we need to understand and familiarize ourselves with an app's functionalities to design its features effectively, we must also grasp the advantages and disadvantages of GPT.
More importantly, we need to identify business scenarios where GPT's strengths can be leveraged.
A GPT AI product manager once told me: the most crucial step in designing GPT AI features is "experimentation."
Once the effectiveness of GPT in a specific business scenario is confirmed, the remaining tasks become relatively straightforward.
Of course, while leveraging GPT's advantages, we must also be mindful of avoiding its disadvantages.
For example, using GPT to generate articles containing SEO keywords works exceptionally well.
However, when generating complex cost analysis reports with GPT, we need to be cautious about its "hallucination" issues.
AI can improve the efficiency of B2B product managers, but it won't make their jobs easier—it might not even prevent the 996 work schedule.
Human desires are endless. Better tools will only make us pursue higher growth. It's said that back in 1930, John Maynard Keynes predicted that due to innovation, by 2030 we would only need to work 15 hours per week.
Now, as we approach 2024, we are still far from the goal of a 15-hour workweek.
In the long run, GPT is indeed very important, even as some have said, it's a 'once-in-a-century opportunity.'
However, at present, GPT's practical applications in the B2B sector remain limited, and it faces many thorny issues (such as the 'black box' problem).
Therefore, the game of GPT has just begun, and there's no need for excessive anxiety.
In fact, for most people, leveraging existing strengths to integrate 'traditional industries + AI' is often more effective than attempting to 'change lanes to overtake.'
Some students, eager to change their fate quickly, readily abandon their original career paths and fully transition into AI products.
There are certainly opportunities for success, but the probability of failure is greater.
Therefore, it's best to consider: if the "career switch" fails, what is your fallback plan?
In fact, to work on AI products, it's not necessary to first become a full-time AI product manager.
In the context of increasingly powerful GPT models and decreasing algorithmic requirements, starting from familiar business scenarios, experimenting more, and finding opportunities for GPT to add value in business operations might actually be a shortcut to becoming an AI product manager.
Remember: No matter how technology evolves, the core of a good product remains differentiated positioning.
And the core of a good product manager is still the ability to identify opportunities.