While explosive growth in artificial intelligence (AI) is augmenting capacities in several sectors, there are also concerns over how it can affect humans. Firms have invested heavily in AI, leaving economists striving to understand the impact on the labour market and driving fears among the wider public for the future of their jobs. The rapid adoption of AI so far is creating and not destroying jobs, especially for the young and highly-skilled, but could reduce wages, research published last year by the European Central Bank has shown.
After ChatGPT made waves all over the world for its surprising generative AI capacity, a US-based company called Cognition has announced the launch of a new AI tool called Devin which it claims to be the world’s first fully autonomous AI software engineer which can write code with command prompts. It has triggered fears among the software community about its possible impact on tech jobs.
Empower Your Corporate Journey with Strategic Skill Courses
| Offering College | Course | Website |
|---|---|---|
| IIM Lucknow | IIML Chief Executive Officer Programme | Visit |
| Indian School of Business | ISB Chief Technology Officer | Visit |
| IIM Lucknow | IIML Chief Operations Officer Programme | Visit |
What is Devin and what it does
As per Cognition, Devin is a tireless, skilled teammate, equally ready to build alongside you or independently complete tasks for you to review. With Devin, engineers can focus on more interesting problems and engineering teams can strive for more ambitious goals.
Devin can plan and execute complex engineering tasks requiring thousands of decisions. It can recall relevant context at every step, learn over time, and fix mistakes.
Cognition has equipped Devin with common developer tools including the shell, code editor, and browser within a sandboxed compute environment — everything a human would need to do their work. Devin has the ability to actively collaborate with the user. It reports on its progress in real time, accepts feedback, and works together with the user through design choices as needed. Devin can learn how to use unfamiliar technologies; build and deploy apps end-to-end; autonomously find and fix bugs in codebases; and train and finetune its own AI models.
Devin correctly resolves 13.86% of the issues end-to-end, far exceeding the previous state-of-the-art of 1.96%. Even when given the exact files to edit, the best previous models can only resolve 4.80% of issues. Cognition tried giving Devin real jobs on Upwork and it could do those too. At present, Devin AI is in the beta testing phase and available to select users in limited access and that too by request. You can request Devin AI access by filling out a form available on their official website.How will Devin impact software jobs?
Devin’s capabilities have raised concerns over its impact on software jobs. Will it prove to be a job-killer as much of AI is being seen, or a blessing for techies who will benefit from it? Cognition presents Devin as a smart assistant that makes the job of software engineers easier and thus allows them to focus on higher-level skills.
Software programming was getting impacted with generative AI tools like GitHub Copilot, but Cognition’s Devin has taken this to another level, Jaspreet Bindra, MD & founder of The Tech Whisperer, has told TOI. “It has seemingly groundbreaking capabilities in transforming software development. It can handle some development projects independently, from writing code to fixing bugs and executing tasks, therefore mimicking a full-fledged AI worker rather than just a coding assistant,” he said.
Its reported effectiveness in software engineering, Jaspreet says, is notable because it can rapidly learn and utilise new technologies, build applications from scratch, identify and rectify bugs, contribute to production repositories, and autonomously train AI models. “This ability to handle complexity is creating a nervous and excited buzz amongst the fraternity,” he says.
However, Devin is being seen mostly as an assistant rather than a competitor. Abhimanyu Saxena, co-founder of Scaler & InterviewBit, has told TOI that software engineers need to see these tools as enablers and quickly build expertise in using them efficiently rather than seeing them as competitors. “It is most likely to be a developer companion and may also enable a lot of non-technical people to easily build applications,” he says.
Coding, Devin’s core capability, is just one part of software development, and that’s why it can’t replace software engineers. Heena Kothari, senior director of engineering and product development at Exotel, has told TOI that Devin represents a big shift in how software is made, and that software development isn’t just about writing code or testing it anymore. “While coding is important, there’s a lot more to it, like planning how the software will work, making sure it fits with other software, and understanding how it’s used in different ways.”
For large enterprise software, Heena says, coding only comprises 40% of the whole software development process. “The rest involves designing the software, making it work with other software, and understanding how people will use it. That’s why Devin could be really helpful for simpler or medium-complexity software projects. It could let engineers focus on solving bigger problems instead of spending too much time on routine tasks.”
Despite its amazing capabilities, Devin may not pose any threat to techies at present but development of generative AI will remain a cause of concern on the jobs front in various sectors, though AI has in fact led to creation of more jobs. The research published by the European Central Bank, cited earlier in this article, is in contrast to previous technology waves, when computerisation decreased the relative share of employment of medium-skilled workers. In a sample of 16 European countries, the employment share of sectors exposed to AI increased, with low and medium-skill jobs largely unaffected and highly-skilled positions getting the biggest boost, a Research Bulletin published by the ECB said.
However, the research says, these results do not amount to an acquittal. “AI-enabled technologies continue to be developed and adopted. Most of their impact on employment and wages – and therefore on growth and equality – has yet to be seen.”
(With inputs from TOI)
