Welcome to Software 3.0

Originally published on July 2023 and written by John White.

The way we write software is changing… In fact, soon you won’t have to write it all. At least, that’s the view of Nvidia’s CEO Jensen Huang. If true, the emerging software landscape may not require many of the skills that developers have built their careers on.

Python, JavaScript, C++. Is Huang right when he says the differences between these languages are about to become irrelevant?

“This computer doesn’t care how you program it, it will try to understand what you mean, because it has this incredible, large language model capability. We have closed the digital divide. Everyone is a programmer. Now, you just have to say something to the computer.”

Jebsen Huang,
Nvidia CEO

More transformative, however, is the fact that in the future, supercomputers are unlikely to even be needed to open up programming to all the non-coders out there. LLM platforms, like ChatGPT, are making software development more accessible than ever. Journalist and US policy advisor David Gewirtz showed exactly what this means in practice.

For comparison, Gewirtz revealed that it probably would have taken him “a few days” to write the WordPress plug-in, and this is coming from an experienced coder. Gewirtz has a degree in computer science and has given lectures online, sponsored by the likes of Google, Dell, IBM, and SAP. As such, it’s hardly surprising that he admitted the experiment had left him “more than a little freaked out.”

“As an experiment, I asked ChatGPT to write a plugin that could save my wife some time with managing her website. I wrote a short description and ChatGPT wrote the whole thing: user interface, logic, and all. In less than five minutes.”

David Gewirtz,
US policy advisor & computer scientist.

So is the rise of generative AI democratising coding or killing it dead?

Back in February, Matt Welsh, a former lead engineer at both Apple and Google, made his thoughts pretty clear during a presentation for ACM Magazines: “Computer Science is Doomed.” In Welsh’s view, the essence of computer science – translating ideas into programs – will soon be irrelevant. After his dramatic opening, however, he did admit that the field has evolved consistently over the decades already, so perhaps it would be fairer to say that computer science (as we know it) is doomed.

Before the emergence of AI-generated code, individuals generally had to spend years completing a computer science degree and then get to grips with the needed design and functionality for any given project. That’s before programmers even got to decide which language to use.

Each has its own strengths and weaknesses, so while JavaScript and Python may be two of the most popular coding languages today, developers still had to consider others like C++, Ruby, Scala or any of the many other scripts. Now, if Gewirtz’s example is anything to go by, the outputs of different languages can all be trialled by simply altering the relevant AI prompt.

Are software developers going to be redundant in the future?

It’s worth remembering that this isn’t the first time that computer programmers have been told to start looking for another job.

The low code/no-code revolution has already touched many businesses and means that they can create bespoke apps simply by using a drag-and-drop interface.

Gartner predicts that by 2025 70% of new applications created by enterprises will be built using low-code/no-code. It’s worth keeping in mind, though, that no code doesn’t mean no skill.

“We don’t see [no-code/low code] replacing the need for development programming engineers, and we don’t see these tools replacing business subject matter expertise. These are tools to draw higher efficiency and productivity from data engineers, designers and the business people who want to contribute to these roles.

We are seeing a blending of teams, so it’s not just designers, architects and database administrators – now you see sales managers, marketing and human resources working together once they learn the low-code and no-code tools.”

Jason Wong,
Gartner Analyst

Wong’s reluctance to get carried away could just as easily be applied to AI-generated code. The query, “Will AI replace software developers?” has crept up the Google rankings in the last few weeks, highlighting the anxiety around the subject. The US Bureau of Labor Statistics also predicts that the number of programmers, in a narrow sense, will decline by 10% between 2021 and 2031. But many other digital roles will be created by the opening up of the coding ecosystem.

“That doesn’t mean the field of software development won’t evolve. Akin to how carpenters improve their efficiency with better tools, software developers now have new tools to work with.

Further, as anyone who’s ever bought a highly sought new piece of technology has probably found, even the sleekest tools require maintenance, troubleshooting and upgrading. This remains true for solutions like ChatGPT, which still requires expertise to run.”

Anthony Hughes,
Co-Founder of coding bootcamp Tech Elevator

For all the concern about job cuts – there were more than 200,000 in the tech sector across 2022 – coding is likely to represent a solid career choice for the foreseeable future. It is still ranked among US News and World Report’s 100 Best Jobs” list, and Microsoft CEO Satya Nadella, speaking earlier this year at a World Economic Forum event, suggested that, far from increasing the number of unemployed coders, AI could actually help plug the tech skills gap by further opening up software development to non-coders. Plus, AI can’t replace the full gamut of skills held by coders – not yet, anyway.

“Every software system has additional considerations – like logical and physical system architecture, data modeling, build and deployment engineering, and maintenance and management activity – that still appear to be well beyond current generative AI capabilities.”

Dr James Fairweather,
Chief Innovation Officer, Pitney Bowes

Rather than viewing AI coding as representative of some sudden break with the past – a sea change that app developers everywhere should fear – perhaps it would be more accurate to view it in the context of coding’s evolution. Just as no-code platforms changed software development without eliminating the software developer, the same is likely to be true of AI-generated code.

This is the age of “Software 3.0”. An age where software development is accessible to everyone. Programmers will still provide instructions for AI coding platforms to work with, but the line-by-line script writing will be taken out of their hands. As famous computer scientist Stephen Wolfram claims, it might be the case that app development depends more on “prompt engineers” than coders in the future. In any case, programmers are likely to be essential for problem-solving, verifying scripts and maintenance – regardless of who or what writes a program’s script.

Is AI the catalyst for the next massive technological convergence?

If programmers need not be overly concerned about the rise of AI coding just yet, maybe investors should be. In the future, start-ups may not be quite so hungry for huge injections of venture capital funding if the resources they need to create an e-commerce app, a mobile site, or any other bit of software are reduced to almost nothing as a result of AI.

Bootstrapping may become increasingly common for new businesses instead. The substantial decline in global VC funding witnessed in April – a fall of 56% compared to 12 months previous – may be a sign of things to come.

However, the best tech investors are the most attuned to pivoting when technological and economic shifts occur, and this seems to be the case considering the increasing amount of VC funding already going towards generative AI platforms themselves.

One area investors are increasingly honing in on is the looming, massive convergence of many of the new technologies we’ve seen emerge in recent years…

AI-generated code will be transformative – but not only for software development itself. It has the potential to impact every sector, from manufacturing to healthcare. This is giving rise to new investment theories trying to identify how combinations of megatrends such as Bitcoin, 3D printing and AI could unlock the next phase of exponential growth.

ARK Invest’s recent Big Ideas 2023 report boldly states that exponential growth can only be found through the convergence of five innovation platforms: Public Blockchains, Multiomic Sequencing, Energy Storage, Robotics and AI. What’s fascinating is that many of these platforms straddle both the digital and physical worlds, combining both software and hardware.

Perhaps, therefore, developers won’t be eliminated at all… We’ll just all increasingly play a role in the next evolution of technology.