AI Mastery Without Coding Knowledge - Tackling Intricate Systems Through Varied Education Strategies

AI Mastery Without Coding Knowledge - Tackling Intricate Systems Through Varied Education Strategies | ZDNet

gorodenkoff/Getty Images

Some top universities are pitching professional development programs for artificial intelligence that seem to require very little of a programming or development background. The University of Pennsylvania, for example, offers a 24-week boot camp that states, “no previous programming experience required.” Not to be outdone, the Massachusetts Institute of Technology offers a 12-week course where you can learn to build AI solutions with no-code software.

Also: Want to work in AI? How to pivot your career in 5 steps

Reading these, one can be forgiven for thinking it’s possible to become an AI master with little or no software development experience. But is that the case? Industry leaders suggest that there are still a great deal of technical chops required to build a well-functioning AI system, but add that strictly technical skills are just part of the equation.

“I would highly caution anyone from thinking that they don’t need to learn basic coding and data analysis skills just because AI can also perform them,” says Dr. Robert Blumofe , CTO of Akamai Technologies. “Not only is it a dangerous mindset that could lead to disregarding all foundational skills as long as they can be done by AI, but you won’t be able to perform quality assurance tasks on AI-generated content.”

Newsletters

ZDNET Tech Today

ZDNET’s Tech Today newsletter is a daily briefing of the newest, most talked about stories, five days a week.

See all

While AI is here to stay, “it, and large language models (LLMs) in particular, have serious limitations,” says Blumofe. “Mainly, LLMs still require human oversight, understanding, and intervention to be used safely.”

“Core technical skills like programming, data science, data management, and data protection will continue to be essential,” says Charman Hayes , executive VP of technology, people, and capability at Mastercard. “At the same time, technologists have a responsibility to understand the evolving legal and regulatory landscape around AI, and this will be critical to ensuring they are using their technical skill sets responsibly.”

Perhaps Ethan Mollick (professor at the University of Pennsylvania, by the way) provides an apt description of technologists’ roles by comparing AI to a “jagged frontier.” This frontier is “where AI excels in some areas while struggling in others, requiring professionals to discern when to compensate for its weaknesses,” relates Cal Al-Dhubaib , CEO of Pandata.

Also: How renaissance technologists are connecting the dots between AI and business

While AI-assisted coding is on the rise, “I don’t see this taking away net jobs from programmers,” says Al-Dhubaib. “However, it does significantly reduce the time it will take to build code and perform data analysis. “I foresee coders spending more time on strategy and orchestrating complex systems, with the expectation to deliver higher value work.”

Still, high-level courses such as those offered by universities may help tech professionals better understand the depth of AI’s impact on their businesses. “Don’t narrow yourself to just deep learning,” Blumofe advises. “Study the full breadth of AI and the foundational technology that underpins it.”

Also: Have 10 hours? IBM will train you in AI fundamentals - for free

“Instead of focusing on just a single set of skills, which could quickly become outdated, technologists should frequently engage in targeted bootcamps and certifications that let them evolve their skills with the needs of the markets,” says Hayes. “Employers should invest in bite-sized real-time learning that employees can do on the job for constant upskilling in compressed amounts of time. For example, Mastercard’s internal opportunity network, Unlocked, helps connect employees to projects, positions, mentorship, and volunteer programs to develop new skills and gain exposure to the broader organization.”

LLMs as a component of AI may have a limited shelf life, Blumofe suggests. “LLMs are amazing and very good at certain tasks, but they also have serious limitations that I expect will become increasingly apparent as people get more experience using them,” he predicts. “I don’t think the next big thing in AI will be a bigger LLM. Rather, it will be something new that either replaces LLMs or relegates them to a narrower role. If you have a basic understanding of how it all works, you will be ready for whatever comes next.”

Also: If AI is the future of your business, should the CIO be the one in control?

As has always been the case with many complex systems, “unintended consequences plague the world of AI and machine learning,” says Al-Dhubaib. “Many of the controversial cases in the news are a result of AI breaking in weird ways unintended by the developers. As AI solutions get more sophisticated, and the data we use with them gets more complex, there are more ways for these models to break. As far as talent is concerned, the need to oversee and validate the safety and efficacy of AI solutions is only going to increase in importance.”

“AI isn’t just creating jobs for data scientists; it’s powering a whole new ecosystem with its own set of needs and opportunities,” says Hayes. “For example, as generative AI takes on the role of information “synthesizer,” certain job requirements may shift, and more time will be available for strategic and consultative work. Some new jobs could focus on oversight (e.g., chatbot manager), as well as interpretation and validation to ensure output accuracy and utility. Other jobs could optimize inputs for a company. These ‘prompt engineers’ will likely continue to grow in demand.”

Examples of roles at Mastercard that integrate and maximize the potential of AI include roles in the areas of “AI governance and AI strategy, as well as AI product management and engineering,” Hayes relates. “Other jobs we see evolving to be more productive and effective include software developers and marketers.”

Also: I confused Google’s most advanced AI - but don’t laugh because programming is hard

Certain evergreen skills will still be in demand for the foreseeable future, says Blumofe. Such skills include “AI algorithms, discrete math, probability, and statistics. If you study those things, your skill set and knowledge will continue to be in demand, no matter what new technologies emerge in the future.”

“I also can’t stress enough how important soft skills are for successful technology careers,” he adds. “Communication, critical thinking, and collaboration are distinctly human skills that can’t be replicated by AI tools.”