Will AI take over all coding?
During last year’s Abundance Summit, Emad Mostaque, CEO of Stability AI, made the statement that we would have “no more humans coding in 5 years.”
Should we embrace this as inevitable and tell our kids they no longer need to learn to code?
There’s strong evidence that AI has already surpassed the ability of human coders, let’s look at three datapoints:
1. In early 2023, OpenAI’s ChatGPT passed Google’s exam for high-level software developers.
2. Later in 2023, GitHub reported that 46% of code across all programming languages is built using Copilot, the company’s AI-powered developer tool.
3. Finally, DeepMind's AlphaCode in its debut outperformed human programmers. When pitted against over 5,000 human participants, the AI beat 45% of expert programmers.
Given that all these developments took place within the first year of ChatGPT’s release, what is likely to happen over the next two or three years as the tech advances even further?
Will AI eliminate the need for human programmers altogether later this decade?
Or, perhaps, rather than eliminate coders, will generative AI allow any and all of us to become coders?
In today’s blog, I want to paint a more hopeful and compelling picture of the future—one that flips our perspective from scarcity to abundance. A future in which more people than ever will be able to leverage the power of coding to solve important problems and uplift humanity.
Let’s dive in…
NOTE: At next month's 2024 Abundance Summit, we’ll have Nat Friedman (Former CEO, GitHub); Mustafa Suleyman (Co-Founder, DeepMind; CEO, Inflection AI); Emad Mustaque (CEO, Stability AI); Eric Schmidt (Former CEO & Chairman, Google); Ray Kurzweil (Google) and many more AI leaders discussing this topic of “AI and coding” and its ability to turn us all into coders in the near future.
AI is Democratizing Coding
In a future where generative AI is doing the coding, anyone who can simply express what they want in natural language (for example, in English), will be able to use AI to convert their desires into code. As NVIDIA CEO Jensen Huang noted during a 2023 earnings call:
"We've democratized computer programming for everyone … who could explain in human language a particular task to be performed."
In this fashion, doctors, lawyers, or kids will code.
By eliminating barriers that once blocked creativity, anyone can now build systems that solve problems and create value for society.
The platforms enabling this revolution are typically referred to as “no-code” and “low-code,” empowering individuals with little to no programming knowledge to develop applications swiftly and economically.
No-code platforms, characterized by a user-friendly interface, facilitate rapid application development for business employees who have understanding in domain-specific areas but limited coding skills, effectively bridging the gap between business requirements and software solutions.
On the other hand, low-code platforms still demand a rudimentary understanding of coding, offering a higher degree of customization and integration capabilities, thus finding preference among IT professionals for more complex tasks. This approach provides a robust tool in the hands of “citizen developers” to create functional applications for back-office apps, web applications, and business automation functions.
But in this new environment, does it still make sense to learn how to code? Should your kids continue to learn Python or another programming language?
While you’re first reaction may be to say “No,” Steve Brown, my Chief AI Officer, has a different opinion:
“Coding is not about a particular computer language or even about writing programs per se. It’s about cultivating a mindset of computational thinking: enhancing your ability to break down complex problems into manageable components, devising logical solutions, and thinking critically.”
This skill will become increasingly important.
While it is true that AI has enabled machines to speak English, if you really want to collaborate with AI and harness its power, learning the native language of AI will give you a distinct advantage.
It’s how you go from a “naive end-user” to an actual creative partner, problem solver, and critical thinker.
Humanity’s Best “Coders” Will be Hybrids
Technology has always allowed individuals to do more, faster. Robotic farm equipment has increased the output of a farmhand by 1,000-fold, while computers have empowered investors, scientists, and digital artists by orders of magnitude.
Now AI, in a somewhat recursive manner, is enabling our best programmers to amplify their skills and programming prowess 100-fold.
AI-enabled programming is a superpower for both the novice and the experienced coder.
AI tools such as Replit and Github’s Copilot are helping developers automate redundant workflows, learn faster, work more efficiently, and scale their productivity.
For example, researchers at Microsoft have found that software developers using AI assistants completed tasks 55% faster than those not using AI assistants. And an MIT study showed that the top 5% of programmers performed orders of magnitude better while partnering with AI.
Now and for the near future, the best coders will be hybrids: humans working with and amplified by AIs.
Why This Matters
By democratizing humanity’s ability to code and by magnifying the abilities of our best coders by 100-fold using AI, we are super-charging our future.
At the same time, AI is also learning how to code itself and improve its own performance and capabilities. Without question, we are accelerating the rate of technological advancement.
While this may scare many, it’s also important to recognize that these improved tools are the superpowers that will enable entrepreneurs to address and slay many of humanity’s grand challenges.
It’s also worth pointing out that these tools are enabling individuals and small teams to take on challenges that were previously only addressable by governments or large corporations.
We are effectively democratizing the ability to solve our biggest problems.
In the next blog in this Age of Abundance series, we’ll explore how AI and AI-human collaboration will transform another industry ripe for disruption: healthcare.