The growing presence of artificial intelligence (AI) in software development has sparked intense debate among tech leaders. While some predict AI will take over programming entirely, IBM CEO Arvind Krishna believes otherwise. Speaking at the SXSW conference, he emphasized that AI is a tool meant to boost productivity rather than replace human programmers.
AI as an Assistant, Not a Replacement
Krishna estimates that AI could handle around 20–30% of coding tasks, but he doubts its ability to manage more complex programming work. While AI can streamline repetitive tasks, he does not see it eliminating the need for skilled developers.
His perspective differs from some industry leaders who foresee a far more significant role for AI. Dario Amodei, CEO of Anthropic, predicts that AI could generate up to 90% of code within months. Similarly, Salesforce CEO Marc Benioff has suggested that AI-driven improvements might make hiring traditional software engineers unnecessary by 2025.
The Value of Human Expertise
Even as he acknowledges AI’s potential, Benioff emphasizes the importance of human skill. Salesforce is already reskilling employees to work alongside AI tools rather than be replaced by them. Krishna shares this viewpoint, arguing that AI will make developers more efficient rather than obsolete.
“If you can produce 30% more code with the same number of people, are you going to get more code written or less?” he asked. In his view, increased productivity will allow companies to innovate faster, expand their market share, and develop new products.
IBM’s AI Strategy
IBM has long been a leader in AI-driven solutions, but the company sees AI as an enhancement rather than a substitute for programmers. While Krishna has previously noted that IBM might pause hiring for certain roles that AI can automate, he stresses that AI’s role in creative and technical fields will be to assist rather than replace workers.
He draws comparisons to past concerns about technological advancements, noting that calculators did not replace mathematicians, nor did Photoshop make artists irrelevant. AI, he argues, is just another tool that will help workers produce higher-quality results. “If the quality that everybody produces becomes better using these tools, then even for the consumer, now you’re consuming better-quality [products],” he said.
AI’s Energy Consumption and Future Efficiency
Looking ahead, Krishna predicts AI will become much more energy-efficient. He pointed to emerging technologies from companies like Chinese AI startup DeepSeek, suggesting AI could eventually use “less than one percent of the energy it’s using today.” If these improvements come to fruition, AI will become even more accessible and cost-effective.
Despite AI’s rapid progress, Krishna is skeptical about its ability to drive groundbreaking scientific discoveries. He believes that AI primarily learns from existing knowledge rather than creating new insights. Instead, he sees quantum computing—a field where IBM is heavily invested—as the key to advancing scientific research.
AI vs. Quantum Computing in Innovation
“AI is learning from already-produced knowledge, literature, graphics, and so on,” Krishna explained. “It is not trying to figure out what is going to come next.”
His stance contrasts sharply with that of OpenAI CEO Sam Altman, who has suggested that “superintelligent” AI could soon emerge and accelerate major breakthroughs. Altman and other AI advocates argue that AI could eventually help uncover new discoveries by identifying patterns that humans might miss.
