Why AI Didn't Fix Knowledge

2025 predictions — graded AI-powered knowledge

Bob's 2025 prediction: AI would dramatically improve knowledge in contact centers. Result: Early but mostly wrong.

The technology moved, but the data did not. Knowledge bases were too fragmented, too dirty, and too poorly governed for AI to meaningfully improve frontline work. The industry instead spent another year chasing bots, automation, and surface-level "AI assistants."

Grade: C+

The failure was not AI. It was the state of enterprise knowledge.

Remote work reversal

Bob's 2025 prediction: Work-from-home would shrink and revert toward pre-COVID norms. Result: Correct.

Remote and hybrid work has fallen to within five percentage points of pre-COVID levels. Companies quietly reversed course not because it helped customers or employees, but because leadership never learned how to manage distributed teams.

Hybrid was the worst of both worlds: frontline leaders juggling physical rooms, video calls, and dashboards without the training or structure to do any of it well.

Grade: A

Why remote work collapsed

The reversal was not ideological. It was operational.

Executives defaulted back to what felt controllable: physical presence. Organizations refused to do the hard work of re-engineering leadership, coaching, quality management, and accountability for a distributed workforce. They solved a people problem with proximity.

Amas' prediction for 2026

Voice comes back.

Digital channels absorbed most of the AI hype: chat, bots, messaging, and self-service. But customers never stopped calling. Voice is where frustration spikes, where trust is tested, and where automation breaks down.

Amas' call: 2026 will be the year voice reasserts itself as the center of the customer relationship — and the CCaaS market will look radically different by 2027 because of it.

Bob's prediction for 2026

Data becomes the bottleneck.

AI will only become useful where it has access to clean, structured, reliable data. The industry rushed into AI before fixing the foundations: knowledge, case data, call logs, customer history, and operational context.

2026 will be the year contact centers slow down, audit their data, and rebuild the plumbing that AI actually runs on.

No data. No intelligence.

What the industry is claiming

Analysts and vendors are promising three things for 2026:

• Predictive and proactive service • Agent empowerment through AI • Fewer humans in contact centers

Bob and Amas reject the third and remain skeptical of the first two without structural change. The hype assumes AI will replace labor. Reality says AI will expose how broken the systems around labor really are.

Amas' 2026 wish

Stop calling software "agents."

For twenty years, "agent" meant a human being doing emotional, cognitive, and relational labor. Rebranding bots as agents erases the workforce and confuses accountability.

Language shapes power. That battle matters.

Bob's 2026 wish

Focus on the employee.

AI should not be used to replace people. It should be used to remove friction from their work: searching, documenting, switching systems, hunting for answers. Knowledge was always the real use case. The industry just skipped the hard part.

Core takeaway

2025 proved that AI without data, governance, and human-centered design does not transform anything. It only adds noise.

2026 will reward the companies that stop chasing demos and start rebuilding the foundations: voice, knowledge, data, and frontline enablement.

That is where the real disruption will come from.