TL;DR
Don Murray, CEO of Safe Software, argues AI is a powerful assistant but should not be treated as the final authority. He warns that relying on AI in place of on-the-job experience risks creating expertise gaps when senior staff retire.
What happened
In an interview, Safe Software CEO Don Murray outlined his view that artificial intelligence is becoming indispensable as an assistant but cannot substitute for experienced human judgment. Safe, a company that aggregates and routes data into visualization and analytics tools, sees its pipelines as well suited to AI applications because the technology needs large volumes of data. Murray cautioned against treating AI as an authority, pointing to engineering projects that demand near-perfect accuracy — far beyond the typical 80–90% AI claims — and saying organisations will require human sign-off on critical decisions. He also raised concerns about workforce development: some utilities are using AI to augment long-tenured experts and conclude they can forgo hiring juniors, a practice Murray says will erode the pipeline of future senior staff. At Safe, Murray says AI is used to accelerate learning and productivity among both junior and senior employees rather than replace expertise.
Why it matters
- Maintaining human oversight preserves accountability for high-stakes decisions in engineering, healthcare and infrastructure.
- Relying on AI instead of hiring junior staff risks starving organisations of hands-on experience needed to replace retiring experts.
- AI’s appetite for data increases the importance of data governance, quality and access across organisations.
- Using AI as an assistant can raise productivity and speed up onboarding if combined with testing and expert review.
Key facts
- Don Murray is CEO of Safe Software, a company focused on data collation and routing to visualization tools.
- Murray describes AI as 'indispensable' as an assistant but not an authority.
- Safe’s data pipelines are considered suitable targets for AI because the technology requires large amounts of data.
- Murray cited an engineering customer example where required accuracy is effectively 99.999%, far above common AI accuracy claims.
- He expects agentic systems to involve a 'human in the loop' who reviews and signs off on AI outputs for critical tasks.
- Some utilities are reportedly using AI with long-tenured experts and considering reducing junior hiring, which Murray warns against.
- Safe uses AI internally to help bring both junior and senior staff up to speed and to suggest code libraries or system support approaches.
- Murray compared AI’s permanence to the internet: it may have had a hype cycle but has become a lasting technology.
What to watch next
- Regulatory or organisational rules that mandate human sign-off on AI-generated designs and decisions.
- Hiring patterns at utilities and infrastructure companies—whether junior recruitment declines as AI is adopted.
- not confirmed in the source: broader industry outcomes on whether agentic AI systems will be allowed to operate without human approval.
Quick glossary
- Agentic AI: AI systems that can perform tasks or take actions autonomously, potentially interacting with other systems or users.
- Human-in-the-loop: A workflow design where humans review, validate or intervene in decisions made by automated systems.
- Data collation: The process of gathering, cleaning and combining data from multiple sources for analysis or routing.
- Upskilling: Training employees to improve their skills and adapt to new tools or technologies.
Reader FAQ
Will AI replace senior engineers and experts?
According to Don Murray, AI should act as an assistant, not a replacement; organisations will still require human authority and sign-off.
Can AI be trusted to sign off on critical engineering work?
Murray says organisations demanding extremely high accuracy will not allow AI to be the final authority and will require human sign-off.
Is using AI a reason to stop hiring junior staff?
The source reports some companies are considering that approach, but Murray warns it would deprive organisations of the experiential training needed for future experts.
Is AI a passing bubble?
Murray acknowledged the possibility of a bubble but stated his view that AI is here to stay; broader industry permanence is not independently confirmed in the source.
AI + ML Safe CEO: AI is an assistant, not a replacement There is no automated substitute for experienced staff, and 'if there's one thing AI has a never-ending thirst…
Sources
- Safe CEO: AI is an assistant, not a replacement
- AI Is An Assistant, Not An Authority
- CEO Don Murray says Safe Software's new features will …
Related posts
- Developer alleges fees and hidden algorithms prioritize profit
- AI Futures Model December 2025 Update: New Timelines and Takeoff Forecasts
- Straussian Memes: How Multi-Level Messaging Enables Mass Persuasion