Overview of IBM’s $10 billion bet on what comes after AI
This episode features a wide-ranging conversation with IBM CEO Arvind Krishna about IBM’s role in the AI era, why he thinks foundation models will become commodities, how enterprises should actually deploy AI at scale, and why IBM is making a major long-term bet on quantum computing. Krishna’s core message: the AI revolution is real, but most companies are still in the “day zero” phase, and the biggest risk is not moving fast enough.
Key Themes and Takeaways
-
IBM is not trying to win the foundation-model race
- Krishna says IBM is not aiming to be OpenAI, Anthropic, Google, or Microsoft.
- Instead, IBM is focused on helping enterprises use AI safely, efficiently, and in the right workload.
- He predicts foundation models will become commodities over time, with switching costs dropping.
-
AI is powerful, but the economics are tightening
- Krishna argues that current AI systems are often like using an “18-wheeler” for every task—too large and expensive for many use cases.
- He expects businesses to shift toward smaller, cheaper, more fit-for-purpose models, especially as token and GPU costs rise.
- He believes the AI market is heading toward a major correction in the next 12–24 months.
-
Enterprises should stop experimenting and start scaling
- Krishna says AI is at “day zero”: the time to move from pilot projects to real deployment.
- His advice is to choose 3–5 concrete use cases, scale them, and learn from implementation.
- He thinks companies should prioritize domain experts who are curious and adaptable, not AI PhDs, for deployment work.
IBM’s AI Strategy
What IBM thinks it’s best at
- Client intimacy and trust
- Enterprise-grade security and handling of IP/data
- Deep technical expertise in specific domains
- Hybrid cloud + AI for regulated and security-sensitive workloads
Why Watson taught IBM a lesson
Krishna was candid about IBM’s past missteps with Watson:
- IBM moved too quickly from building blocks to a vertical application.
- It chose healthcare, a domain where IBM lacked enough direct selling and regulatory expertise.
- The strategic mistake was trying to create a monolithic solution instead of building reusable infrastructure.
How IBM uses AI internally
- IBM says its software developers are now about 40% more productive than two years ago.
- That productivity gain helped IBM unlock roughly $4.5 billion in efficiency.
- Krishna emphasized that AI savings often take time: early on, companies may spend more than they save before returns compound.
AI, Jobs, and the Workplace
Krishna’s view on job displacement
- He expects some back-office roles to shrink, especially in functions like:
- compliance
- accounts payable
- procurement
- He suggested that as much as 30% of headcount in some operational areas may no longer be needed over time.
But he also expects net job growth
- IBM has tripled entry-level college hiring this year.
- His logic: if AI lowers the cost of software and operations, companies can build more products and create more value, which can mean more hiring elsewhere.
Leadership responsibility
- Krishna believes business leaders should:
- upskill
- re-skill
- create opportunities for workers to transition
- But he also said employees must be willing to adapt; companies can offer the path, but not force the change.
Cybersecurity: AI Cuts Both Ways
Why AI changes the threat landscape
- AI makes it easier for more people to find vulnerabilities in code.
- Krishna said what used to require a small number of experts can now be done by people with average skill.
- That means the attack surface expands.
IBM’s response
- IBM launched a $5 billion initiative to identify and fix AI-related vulnerabilities in open-source software.
- The company aims to create a kind of clearinghouse for vulnerability disclosure and patching.
- The goal is both defensive and commercial: improve security, but also offer the service at a fair price.
Broader warning
- Krishna warned that smaller organizations and previously non-targeted systems will become more vulnerable.
- His blunt advice: if you think you’re protected, it’s only a matter of time before you get targeted.
IBM’s Big Quantum Computing Bet
Why quantum now
- IBM is investing heavily in quantum because Krishna believes it can become a real commercial advantage within a few years.
- He framed quantum as the next platform shift after CPUs and GPUs.
Why the progress matters
Krishna highlighted a rapid jump in quantum capability:
- from simulating a 5-atom molecule
- to 300 atoms
- to 12,000 atoms, which enters the realm of protein-scale problems
Why this is important
- Quantum could unlock breakthroughs in:
- biology
- chemistry
- drug discovery
- fluid dynamics
- aerodynamics
- IBM’s belief is that quantum will soon solve problems that are currently impractical or too slow for classical computing.
Krishna’s Advice on Risk and Innovation
The biggest risk is taking no risk
Krishna’s strongest leadership message was:
- The most risky route is taking zero risk.
Why conservative companies decline
- If a company avoids innovation:
- profits get squeezed
- competitors copy the most profitable parts
- the business slowly becomes less relevant
- In his view, caution can create a declining profit pool that eventually leads to collapse.
How to manage risk well
- Not every experiment needs to win.
- Leaders should create a culture where teams can try things with a reasonable probability of success, not demand perfection.
- His advice: accept that some initiatives will fail, but enough should work to fund the next wave.
IBM in 2030
Krishna said IBM wants to be known for:
- bringing innovation to clients in a way they can actually consume
- being a leader in AI deployment, not just AI invention
- becoming more software-centric
He noted IBM has already shifted from about 20% software in 2019 to around 45% today, and expects that trend to continue.
Notable Insights
- “Foundation models are going to become commodities.”
- “We’re using the 18-wheeler for everything.”
- “This is day zero.”
- “The biggest risk is taking no risk.”
- AI deployment is less about needing PhD-level expertise and more about finding motivated domain experts.
- The next wave of enterprise AI will likely be about economics, specialization, and scale, not just model size.
Bottom Line
Arvind Krishna’s view is that AI is transforming business, but the winners won’t be the companies that merely adopt the biggest models first. They’ll be the ones that:
- pick the right use cases,
- manage cost carefully,
- use AI in fit-for-purpose ways,
- prepare for cybersecurity changes,
- and invest ahead in the next platform shift, especially quantum computing.
IBM’s strategy is essentially to be the company that helps enterprises navigate what comes after the initial AI hype—when the economics, infrastructure, and real operational value become what matter most.