The rapid ascent of AI in the tech realm has often been met with a mixture of awe and skepticism. One area significantly impacted by these advancements is education, particularly coding bootcamps. The emergence of ChatGPT and other Large Language Models (LLMs) represents a profound shift in how we approach coding and, by extension, how we teach it.
Training developers in an era of intelligent machines
Coding bootcamps rose to prominence by promising -and, to a large extent, deliverin- fast, intensive courses that equip students with employable tech skills. The appearance of LLMs capable of generating code, error-checking, and even offering coding advice represents a fundamental shift in this paradigm. On one hand, ChatGPT can serve as an ever-available mentor, guiding students through complex problems and offering instant feedback. On the other, there’s a looming shadow—could these models outpace and potentially replace the very programmers they’re helping to train?
Thus, the ability of ChatGPT and similar large language models (LLMs) to write code has sparked a discussion about the future of human programmers. Many believe that LLMs can handle repetitive or boilerplate coding tasks, reducing the need for entry-level coders. With the ability to understand vast amounts of code, LLMs can assist in finding errors and suggesting bug fixes, reducing the time developers spend on these tasks. Given the right instructions, LLMs can code rapidly, potentially outpacing human programmers for specific tasks.
An alternative view, on the other hand, sees the net balance of the impact of LLMs as producing a growing demand for human programmers. According to this view, the code generated by LLMs requires human oversight for quality assurance, ensuring it meets the required standards and serves the intended purpose. In the end, LLMs, like any tool, need human operators. Developers will be required to train, guide, and maintain these models. More broadly, as with any technological advancement, new opportunities arise. Humans will be needed to explore and develop applications leveraging LLMs, potentially leading to new professional domains and specialties.
A clearer answer to these contrasting perspectives will only emerge with time. The consensus presently leans toward the idea that LLMs won’t replace human programmers wholesale but will change the nature of their work. Programmers might spend less time on repetitive tasks and more on complex problem-solving, design, and strategy. Adaptability and continuous learning are becoming even more crucial for developers. In the meantime, there is no question that coding bootcamps need to adapt to the new AI era. They will have to produce very soon a new breed of coders—ones who not only understand the intricacies of programming but can also guide and collaborate with AI tools to produce optimal results.
From pure coding to supervision. From hard to soft skills
As LLMs increasingly automate repetitive coding tasks, the role of human programmers may evolve. The core of a programmer’s job could shift from actual coding to supervising, guiding, and quality-checking the code produced by AI models. This isn’t isolated to the world of professional programming; coding bootcamps may need to adjust their curriculum to emphasize these supervisory skills. And while LLMs can teach syntax and handle repetitive tasks, they’re yet to master the art of intuition, creativity, and the intricate problem-solving that complex software development demands.
An implication is that he reality of machines taking on more coding tasks underscores the increasing value of human soft skills. The likes of ChatGPT can handle hard syntax, but can they truly brainstorm, collaborate, and deeply understand human-centric problems? Coding bootcamps, in the face of LLM emergence, might find a renewed purpose: to nurture these essential soft skills -such as creativity, teamwork, leadership, conflict resolution, work ethic…-, focusing on architectural understanding and design thinking when it comes to build software.
From challenges to opportunities
Beyond the points raised above, the primary challenge for bootcamps will be to pay tribute to its reputation as nimble and adaptive as far as the curriculum is concerned, by quickly including in their curses everything a programmer needs to know in the area of AI if it is going to be employable. And there are signs that this is already taking place, at least among the market leaders.
Beyond it, founders and managers of bootcamps may found also a whole new set of enticing opportunities open as a consequence of the availability of ChatGPT and similar LLMs. A personalized AI tutor could be assigned to every single student, considerably reinforcing the advice and follow up provided by human instructors. Similarly, ChatGPT could be integrated in lessons, workshops, hackathons or seminars, providing assistance to participants in real time. On the business side of bootcamps, AI may become instrumental in answering convincingly FAQs, post-bootcamp follow up and support, more effective introductory courses pre-bootcamp and other efficiency gains that may result in reduced operational costs.
In the middle of all these possibilities, it will most likely be essential for coding bootcamps not to lose sight of the strengths of their methodological approach to teaching and learning. Generalizing a bit, and allowing for the many variations found in so many bootcamps, such approach usually emphasizes personalized training and mentorship, structured and personalized guidance to student of diverse backgrounds, and a personal touch with a lot of teamwork and human interaction with instructors.
A quick look at bootcamps in Latin America
Leading bootcamps in Latin America are already heeding these messages. A quick survey of four outstanding bootcamps operating -and in some cases even originally founded- in the region indicates that the tools required to generate code using AI -ChatGPT but mostly other specialized resources ready made for coding professionals- are already being included in their courses. The bootcamps have not received direct and strong demand emanating from businesses in this regard, but they perceive a gradual spread of the use of ChatGPT and LLMs models and want to make sure that they graduates are adequately equipped or their employability may soon be at risk.
A particular path most are already following is the development of personalized AI tutors, individualized for each student and able to accompany each and everyone at their own pace, something they see enhancing learning, reinforcing the role of human instructors and tutors, rather than replacing them. Another, equally innovative, is taking advantage new tools for training students that are missing the resources or the disposition to immerse themselves in learning how to code: participants in these innovative courses can develop an app or contribute to the automatization of a process with tools that involve no coding, strictly speaking.
The overall perspective is that the advent of new AI tools will not threaten the mission, core methodological approach or the continuity of coding bootcamps, and no weakening of the demand for developers is in sight in the labor market, exception made of eventual economic downturns. Bootcamps leaders remain optimistic, as long as each bootcamp that wants to endure quickly adapt its curriculum and its methods to the new technological developments. In time, the market will move in the direction of transforming the role of the developer, probably turning it into more strategic or much more efficient, and all that will affect the content of the courses and the types of positions and compensation that programmers will get, but a wholesale replacement of human developers by machines does not seem to be in order, or at least not before several decades in the future.
*The author wants to recognize the contributions made by a selection of coding bootcamp founders and leaders working in Latin America, who answered my questions on issues relevant for this blog. They are Marcelo Ricigliano (4Geeks Academy), Sebas Buffo Sempé (Le Wagon, Latin America), Nayib Abdalá (Make It Real) and Diego Arias Gómez (Desafío Latam).
Read more by the author here.
Leave a Reply