As Artificial Intelligence and Machine Learning continue to disrupt all industries, it's no surprise that 92% of developers are already using AI-powered coding tools in their daily work to improve quality, enhance user experience, streamline processes and reduce cost.
While the role of the developer is evolving, those who can master adaptability, embrace new cutting-edge technologies, and continuously expand their skill set, will reap the benefits of AI and position themselves as invaluable assets in the industry.
For the first time in over three years, Tech in Motion brought together a panel of experts and a live crowd full of excited technologists at Chicago's 1871 to discuss how developers can future-proof their careers by understanding the opportunities available within the exciting world of AI.
The panel included our moderator Ari Kaplan Head of Evangelism @ Databricks; and panelists Tery Lockitski Board Member @ i03 Partners; Helen Sun AI/ML Engineering Leader@ Meta; Asaf Elani Director, Solution Engineering - North America @ C3 AI; and Fouad Bluestone Director of Machine Learning, Vision AI, and Innovation @ Grainger.
Below is a recap of the discussion; the full conversation is available to watch anytime on demand here.
Everybody wants to know, what is Generative AI?
Fouad: We can write books on this question. Every industry will be transformed through AI. In a nutshell, Generative AI is a class of AI and machine learning algorithms and large language models that can generate content as good as human experts, or sometimes better. It’s a big black box that takes input and generates output. Generative AI is trained on large amounts of data and generalized on different tasks of language.
There are plenty of applications. Beyond text and images, there are tremendous applications in medicine and healthcare. Generative AI will help accelerate drugs and new product development, or help doctors to summarize notes and customize reports for patients. In manufacturing, it will help automate and optimize day-to-day operations and processes. In education, it will help educators develop curricula and customize and tailor content. In art and design, it can take some text and generate a piece of art or design for marketing purposes. I could keep going on for an hour about all the use cases.
How is this changing role of the software development process?
Fouad: It’s a paradigm shift, like the development of the Internet in the 90s. Generative AI can generate blocks of code and solve very complex problems to bring solutions to market quickly. For engineers, it’s exciting but also kind of scary. Engineers are concerned that AI may take over their jobs.
However, it will not eliminate jobs, it will repurpose the way we do jobs. In software, there are multiple repetitive tasks like writing stories or blocks of code. AI can do this repetitive work. As a result, engineers will be more productive and efficient in terms of things like putting the code into production.
The future is heading toward more testing. Because with Generative AI, there is potential for more threats and vulnerabilities, we need more test engineers.
Tery: If anything, we are going to need to build up more skill sets. Managers like me will want to get back to being more hands-on. The only way to do that is to automate some of our platforms, so we can get back to doing more innovative and fun things.
Asaf: We are seeing Generative AI affecting all phases of the software development cycle, but you still need the human context. Constraints cannot be solved by just feeding a prompt into Generative AI. They must add context and relevance.
Tery: If we use a lot of generative AI without context, problems arise and we build a lot of technical debt, which management people like me hate.
Helen: You might be thinking how do I ride the change, how do I not let it happen to me? One way is to focus on your core competencies. A while ago, I was hiring a data architect and asked him specific questions. I was not looking for specific technical skills. I bought my business card I asked him to design a data model just on the information from the business card.
His reaction was interesting. He really got into it and was excited when asked about how to do this. He is asking questions and getting curious. He asked me, “What am I designing this database for, what is it going to be used for?” He got hired and turned out to be a great hire.
We talk about the skills that you need. Behavior skills are super important. Collaboration, problem-solving, and working with people. Resolving conflicts effectively and having the ability to debug and navigate through ambiguity. These things will not change regardless of the technology.
Not to say you don’t also need transient skills. Transient skills change all the time. Today’s technology has the lifespan of a banana peel. New features come out all the time and you need to keep learning That will get you in the door. Companies expect you to be productive right away.
Foundational skills are still needed too. It used to be in computer science you needed parallel programming, algorithms, object-oriented programming, functional programming, SQL, background in networking, storage, etc.
With AI, the foundational skills you need are need math, statistics, linear regression, and calculus.
Transient skills get you in the door. Foundational and behavioral skills help you keep growing.
Tery: Showing leadership is not just managing people. How productive can I be and manage tasks and my own time? How productive can I be so I can leverage those foundational skills?
Learn how to manage a project and a budget. No amount of Generative AI is going to replace these soft-touch skill sets.
Fouad: There are some elements of software development that are not replaceable. Such as aligning with business partners to understand key deliverables and requirements. Human intelligence needs to be part of the process to accomplish and present work to business partners and stakeholders.
I strongly believe engineers are not replaceable. Generative AI will make the day-to-day work of engineers more fun, and creative. It’s just another tool in their tool chest.
Now there is a new realm called ML ops, the operationalization of machine learning. How do you respond when the data drifts? How does it affect your models? How do you govern it? Allow transparency to understand data?
Fouad: Prompt engineering will become a very important skill set to have. You need to know how to engineer or prompt models to get the right outcomes. Put this on your resume. “I am a prompt engineer.” It could be generating code. It could be creating a campaign for marketing. This will be a huge opportunity for engineers to understand how to interact with Generative AI.
Even for non-engineers, Generative AI can be useful. Everybody writes emails, right? Learning how to use prompts to use it to edit and improve emails will become a valuable skill to have.
Helen: Excel is a most underrated tool. Super useful. But if we can uplift tools like Excel even further with prompt engineering, I think the power of analysts can be magnified tremendously in this way.
