Recently, a viewpoint has been making the rounds on Wall Street: traditional software is on its decline, large AI models are set to outshine conventional software firms, and those involved in coding, SaaS development, or client-side application creation will soon find themselves obsolete.

In truth, this scenario is just another iteration of the 'wolf is coming' narrative in the AI age: the notion that AI will revolutionize everything overnight. And there will always be individuals eager to engage in such discussions, as they attract significant attention.

However, this narrative oversimplifies a prolonged industrial transformation into a sensational story of 'instant disruption,' which is quite far-fetched.

1. Every technological leap sparks claims of obsolescence.

Looking back over the past few decades, nearly every major technological shift has been accompanied by a similar narrative:

The old is declared dead, and the new is exalted.

When PCs and graphical interfaces first emerged, some dismissed personal computers as mere toys with limited potential.

As PCs gained widespread acceptance, predictions arose that mainframes, data centers, and command-line interfaces would vanish.

What actually happened? PCs not only spread globally, but data centers also thrived, growing in size and significance.

After graphical interfaces became mainstream, command-line interfaces didn't end up in the annals of history; instead, they became the foundation for cloud computing and mobile systems.

Today's most prevalent and efficient 'new interactive' devices still essentially adhere to the principles of command-line interfaces.

Now, consider the retail sector. Many predicted the demise of physical stores last century. For instance, the tech community and Wall Street once viewed Walmart as a 'relic of the past,' believing it couldn't compete with Amazon.

What transpired? Amazon experienced its share of ups and downs and was once derided as an 'internet bubble;' Walmart, on the other hand, embraced an omnichannel strategy, blending online and offline approaches and integrating its logistics.

Decades have passed, and both companies have grown into trillion-dollar giants, neither replacing the other. What has changed in the retail industry are the channels, efficiency, and customer experience; what remains constant is continuous innovation: from small shops to department stores, shopping malls, hypermarkets, online platforms, and now, the integration of online and offline shopping... Factors such as taste, logistics, and product variety have also continuously evolved, shaping the industry's core.

The media landscape is no different. When Netscape first gained popularity, the prevailing belief was that content is king, and new media would render traditional media obsolete. News would cease to exist, music would be free, and professional content would be supplanted by user-generated content (UGC).

Later, with the advent of iPods, smartphones, Netflix, and streaming services, discussions arose about the potential disappearance of cinemas, the closure of TV stations, and the unemployment of professional creators.

Yet, the reality is that news is still thriving, albeit with a new business model combining advertising and subscriptions; music hasn't died either, merely transitioning from CDs to streaming; film and television content has exploded in growth; professional creation hasn't vanished, nor has UGC dominated the scene. Instead, the two have complemented each other, expanding the overall media landscape.

Thus, history across various sectors has consistently shown that claims of 'complete extinction of XX, total replacement by XX' rarely materialize. Humans tend to overestimate short-term changes and underestimate long-term transformations; they relish the idea of 'instant disruption' but are hesitant to accept 'gradual penetration.'

2. Is the notion of 'software is dead' itself a misinterpretation?

Now, in the AI era, the logic and narrative remain strikingly similar.

Large AI models can write code, create interfaces, automate processes, and run intelligent agents, leading some to claim that traditional software is doomed and software companies have no future.

In response, we can present three counterarguments.

1. AI will not replace software; it will only enhance the software ecosystem.

Traditional software serves as a tool, a process, an interface, and a set of logical operations.

Future software in the AI era will build upon this foundation by adding understanding, generation, decision-making, and automation capabilities.

This is analogous to how graphical interfaces didn't eliminate command-line interfaces, e-commerce couldn't wipe out offline retail, and streaming media didn't dismantle content creation. In essence, AI will only add intelligence and convenience to existing software, not replace it.

Moreover, the essence of software is to solve real-world problems through logic and processes. As long as human challenges persist, software will remain relevant, at least not through simple replacement.

2. The demand for software has not yet peaked.

Admittedly, there is already an abundance of software today, but from a broader economic and social perspective, only a fraction of the world has truly been transformed or is participating in the transformation through software.

For example, factories (especially small and medium-sized ones) still rely heavily on workers' experience for many processes; inventory, customer management, and financial collaboration methods in SMEs remain rudimentary; and many scenarios in households, communities, public services, healthcare, education, and transportation haven't even completed basic digitization...

The popularization of AI will not diminish the value of software; instead, it will create an even larger demand gap.

AI's ability to automatically write code doesn't mean there will be less software; rather, software can be produced more rapidly and become more widespread. AI intelligent agents automatically building products doesn't mean developers will become unemployed; instead, ordinary people can quickly develop their own small tools and systems.

3. Specialized skills in niche areas will become increasingly valuable.

Since GPT gained global attention, a familiar anxiety has resurfaced.

AI seems capable of everything, suggesting that lawyers, doctors, designers, consultants, analysts, writers, painters... will all be replaced.

While no one can truly predict how far AI will advance in the future, we certainly shouldn't assert that such replacement will definitely not happen. But for now, it's clear that AI is primarily replacing auxiliary, repetitive, and purely executive tasks, while those with industry knowledge, scenario understanding, human insight, risk assessment, and decision-making abilities are becoming increasingly scarce and valuable.

For instance, AI can quickly draft contracts, but without professional guidance, the precision and relevance of its output will undoubtedly be lacking. Similarly, AI can assist in diagnosis, but without understanding medical history, individual differences, and clinical experience, it cannot directly replace doctors.

It should be emphasized that AI is an amplifier, not a replacer. It enhances efficiency, productivity, and possibilities but does not substitute for industry accumulation, professional judgment, and practical experience grounded in real-world operations.

4. Companies will undergo reshuffling, but not overnight.

It's undeniable that some companies will not survive. Those that cannot keep pace with technology, refuse to transform, cling to outdated models, or rely solely on information asymmetry to generate revenue will gradually be phased out, with or without AI. Players will change, markets will shift, and business models will evolve, but the entire industry will not undergo a sudden transformation; instead, it will be gradually reshaped by the times and technology.

The idea that AI will completely dismantle the software industry within one or two investment cycles defies the laws of social development and technological transformation.

3. Don't Be Misled by 'Endgame Fantasies'

We believe that many people who proclaim 'software is dead' are essentially seeking attention or are obsessed with a theoretical endgame:

AI is omnipotent → code is automatically generated → software runs autonomously → humans no longer need to write software or create products.

But as this report reveals, the real world never jumps directly from point A to point Z. Declaring 'software is dead' today is essentially the same misguided interpretation as declaring 'mainframes are dead,' 'physical stores are dead,' or 'media is dead' in the past: seeing the potential of technology but distorting the pace of industrial change; understanding the capabilities of tools but underestimating the resilience of systems; embracing future visions but forgetting the lessons of history.

AI is not the terminator of software; it is the new starting point for software. A vast, more complex, and more thrilling new era has just begun.

END