--- title: "The physical edge: where do workers fit in the age of AI?" description: With AI on the rise, it’s tempting to ask which jobs it will replace. Here’s why that’s the wrong question, and where AI leaves frontline workers. image: https://blog.zello.com/hubfs/marine-service-lead-directing-by-voice.jpg --- - Why Zello ### Solutions - [Frontline Communication](https://zello.com/product/frontline-communications/) - [Radio Replacement](https://zello.com/solutions/upgrading-to-zello/) - [Dispatch](https://zello.com/product/dispatch-hub/) - [Customer Assistance](https://zello.com/product/kiosk/) - [Management](https://zello.com/product/management/) ### Features - [Channels](https://zello.com/product/features/channels/) - [Message History](https://zello.com/product/features/history/) - [Emergency Alerts](https://zello.com/product/features/alerts/) - [Transcription](https://zello.com/product/features/transcription/) - [Reliability](https://zello.com/product/features/reliability/) - Industries - [Aviation](https://zello.com/industries/aviation-push-to-talk/) - [Retail](https://zello.com/industries/retail-push-to-talk/) - [Transportation](https://zello.com/industries/transportation-push-to-talk/) - [Construction](https://zello.com/industries/construction-push-to-talk/) - [Manufacturing](https://zello.com/industries/manufacturing-push-to-talk/) - [Hospitality](https://zello.com/industries/hospitality-push-to-talk/) - [Emergency Response](https://zello.com/industries/emergency-response/) - Resources - [Customer Stories](https://zello.com/customers/) - [Blog](https://blog.zello.com/) - [Hardware](https://zello.com/hardware/) - [Partner Portal](https://zello.com/partners/) - [Developer Tools](https://zello.com/developers/) - [Support Centers](https://zello.com/support/) - [Trust Center](https://trust.zello.com) - [Downloads](https://zello.com/downloads/) - [Enterprise](https://zello.com/enterprise/) - [Pricing](https://zello.com/pricing/) - - [Contact Sales](https://zello.com/contact/talk-to-sales/) - [Sign In](https://zello.com/work/login/) - [Contact Sales](https://zello.com/contact/talk-to-sales/) - [Sign In](https://zello.com/work/login/) [Try for free](https://zellowork.com/home.htm?user_cohort=master&__hstc=18491174.b7c2db323bb8152d5171aebe5ee59853.1728928254831.1728930098421.1728935329442.3&__hssc=18491174.1.1728935329442&__hsfp=544353888)[Call Sales](tel:+15125804675) - [Try for free](https://zellowork.com/home.htm?user_cohort=master&__hstc=18491174.b7c2db323bb8152d5171aebe5ee59853.1728928254831.1728930098421.1728935329442.3&__hssc=18491174.1.1728935329442&__hsfp=544353888) - [Contact Sales](https://zello.com/contact/talk-to-sales/) - [Sign In](https://zello.com/work/login/) - Solutions - [Frontline Communication](https://zello.com/product/frontline-communications/) - [Radio Replacement](https://zello.com/solutions/upgrading-to-zello/) - [Dispatch](https://zello.com/product/dispatch-hub/) - [Customer Assistance](https://zello.com/product/kiosk/) - [Management](https://zello.com/product/management/) - Features - [Channels](https://zello.com/product/features/channels/) - [Message History](https://zello.com/product/features/history/) - [Emergency Alerts](https://zello.com/product/features/alerts/) - [Transcription](https://zello.com/product/features/transcription/) - [Reliability](https://zello.com/product/features/reliability/) - Industries - [Aviation](https://zello.com/industries/aviation-push-to-talk/) - [Retail](https://zello.com/industries/retail-push-to-talk/) - [Transportation](https://zello.com/industries/transportation-push-to-talk/) - [Construction](https://zello.com/industries/construction-push-to-talk/) - [Manufacturing](https://zello.com/industries/manufacturing-push-to-talk/) - [Hospitality](https://zello.com/industries/hospitality-push-to-talk/) - [Emergency Response](https://zello.com/industries/emergency-response/) - Resources - [Customer Stories](https://zello.com/customers/) - [Blog](https://blog.zello.com/) - [Hardware](https://zello.com/hardware/) - [Partner Portal](https://zello.com/partners/) - [Developer Tools](https://zello.com/developers/) - [Support Centers](https://zello.com/support/) - [Trust Center](https://trust.zello.com) - [Downloads](https://zello.com/downloads/) - [Enterprise](https://zello.com/enterprise/) - [Pricing](https://zello.com/pricing/) ![](https://blog.zello.com/hubfs/marine-service-lead-directing-by-voice.jpg) ### October 16, 2026 #### 5 minute read # The physical edge: where do workers fit in the age of AI? ## With AI on the rise, it’s tempting to ask which jobs it will replace. Here’s why that’s the wrong question, and where AI leaves frontline workers. - [AI](https://blog.zello.com/tag/ai?hsLang=en) - [Frontline communication](https://blog.zello.com/tag/frontline-communication?hsLang=en) - [Frontline OS](https://blog.zello.com/tag/frontline-os?hsLang=en) ##### [![Bren Lawrence](https://blog.zello.com/hubfs/Hubspot%20profile%20icon.png) Bren Lawrence Zello Team](https://blog.zello.com/author/bren-lawrence?hsLang=en) The usual question about AI is how many jobs it will take. Underneath it is a bigger question about what place people will have in the working world. Answering that means establishing where AI cannot go and what it cannot do, a boundary we call the physical edge. Asked in terms of jobs, though, the question is badly formed. A job is too coarse a unit to show where the impact of AI falls. Almost every job is a bundle of different kinds of activity, and automation never takes the whole bundle at once. The question needs a different starting point. Locating the physical edge means setting jobs aside and looking at work. ## **Where automation has its effect** Jobs and work are easy to treat as the same thing, and they are not. People hold jobs and do work. Any one job is a bundle of different work held together by a title. Keeping the two apart is what makes it possible to say where automation has its effect. Take the job of a firefighter, for example. Holding that title means doing all of this work: - Reading fire behavior to decide where to go in - Forcing entry through a locked door - Advancing a hose line into the building - Carrying an unconscious adult down a stairwell - Giving first aid to someone pulled out of the building - Checking and maintaining gear between calls Now consider what it would mean for AI to take the firefighter’s job. Taking the job would mean doing everything on that list, because a job is nothing over and above the work it is made of. Some of that work is within AI's reach. Reading fire behavior means judging from the smoke and heat where a fire will spread next and whether the structure will hold, and AI is already close to making that kind of judgment from camera and sensor data. Carrying an unconscious adult down a stairwell, on the other hand, is something AI cannot do at all, and getting better at reading fire behavior will not change that. ![](https://blog.zello.com/hs-fs/hubfs/image-png-Oct-14-2026-05-52-25-2819-PM.png?width=2800&height=900&name=image-png-Oct-14-2026-05-52-25-2819-PM.png) AI can take on part of the firefighter's job, but not all of it, so the firefighter still has a place. Seeing what AI will change in any job means looking at the work inside it, because the job as a single unit can't show you that. Every job mixes different kinds of work the way the firefighter's does, and sorting the different kinds of work out affords us a better picture of which work AI can take on and which it can't. ## **The six kinds of work** Before we can say how AI affects work, we need a picture of what the kinds of work are. Work can be sorted by asking one question about it: what is the person doing it primarily trying to change or maintain? Asking that question sorts all work into six kinds: - **Physical work** changes matter, energy, or location. Construction, farming, warehousing, maintenance, cleaning, repair. - **Operational work** changes or maintains the state of a machine, vehicle, or live process. Driving, piloting, crane operation, production line control, dispatch. - **Knowledge work** changes information, models, judgments, or decisions. Engineering, accounting, analysis, planning, diagnosis. - **Coordination work** changes the actions of other people. Management, sales, teaching, negotiation, scheduling. - **Care and service work** changes another person's condition or experience. Healthcare, hospitality, childcare, emergency response. - **Creative and experiential work** changes meaning, emotion, attention, or imagination. Art, design, storytelling, performance, entertainment. None of these kinds of work is new, and every one of them goes back at least as far as ancient Rome. What has changed is how much of the workforce each one takes up. Early agricultural societies ran almost entirely on physical work, and in the middle of the nineteenth century around half of American workers were in farming. Today about one percent are, because machines took over most of that physical labor. As people left farming, many of them went into operational work instead, driving trains and running production lines. Computers and telecommunication networks then set off the largest shift yet, moving a large share of the workforce into knowledge work. Machines and then computers each changed the mix of work people do, and AI is changing it again. To see how, we can take each kind of work in turn and ask what AI does to it. ![warehouse-supervisor-talking-hands-free-earpiece](https://blog.zello.com/hs-fs/hubfs/warehouse-supervisor-talking-hands-free-earpiece.jpg?width=670&height=388&name=warehouse-supervisor-talking-hands-free-earpiece.jpg) ## **What AI can and cannot do, by each kind of work** To find the physical edge, we can go through each kind of work one at a time, looking at whether AI affects it and how. **Knowledge work** is the natural place to start, because it is where AI works most directly. Information and judgments are exactly what AI produces, and it has already come a long way at drafting designs, forecasts, diagnoses, and plans. A developer who used to spend the day writing code now spends more of it directing something else that writes code. In large part, what was knowledge work has become coordination work. **Coordination work** has grown as more knowledge work has turned into directing AI. Coordination means directing, and agents can now be directed as well as people. Delegation used to require being a manager, and most people never did much of it. Anyone who hands a task to an AI agent is doing coordination work, so the more jobs use AI that way, the more of them contain coordination work. **Creative and experiential work** is changing in much the same way knowledge work has. AI now sits inside the platforms where design work already happens, so a first draft or a mockup no longer has to come from a designer. What takes skill now is judging what comes out and finishing it. **Operational work** is changing too, as more of it turns into knowledge work. As machines have become more autonomous, operators spend less time working the controls and more time watching for anything out of the ordinary and deciding when to step in. AI is speeding up that shift. AI is changing all four of these kinds of work, so the physical edge has to be somewhere else. ## **What AI can't change** Two kinds of work are left: physical work, and care and service work. Both have to be done in person. Fixing a pipe has to happen where the pipe is. Calming a frightened patient has to happen where the patient is. However much better AI gets, it can't change where the pipe is or where the patient is. Physical, care and service work are where the physical edge is. ## **The physical edge is where the frontline lives** No matter the industry, frontline work sits at the physical edge, made up of some mix of physical, care and service work. Ramp agents turning aircraft spend almost the whole shift on physical work, loading bags and guiding planes to the gate. Store associates split their time between physical work like stocking shelves and service work like helping customers. Ramp agents and store associates do very different jobs, but most of what either of them does has to happen in person, which holds for frontline work generally. Automation is making those workers more important to the businesses they work for. The more of an operation a company automates, the more of it depends on the people at the physical edge. Machines in an automated warehouse still jam and break down, and they still run into things they weren't built for, like a crushed box or a pallet stacked wrong. When that happens, someone has to walk over and deal with it by hand. Every automated process depends on people on the floor to step in when something goes wrong. Yet frontline workers are the people a business is least connected to. Most of the systems a business runs on were built for people at a desk. Schedules, alerts, work orders, and AI tools all sit on screens, and very little of that information reaches a worker whose hands are full. Information coming back from the floor has the same problem. When a worker spots a damaged pallet or a customer who needs help, what they saw usually gets said over a radio call and goes no further, and the systems running the business never learn it happened. ![utility-worker-reporting-by-voice-speaker-mic](https://blog.zello.com/hs-fs/hubfs/utility-worker-reporting-by-voice-speaker-mic.jpg?width=670&height=394&name=utility-worker-reporting-by-voice-speaker-mic.jpg) ## **Why the frontline needs a voice-first Frontline OS** Connecting frontline workers to the business takes a tool that works at the physical edge. Frontline work happens in person, with a worker's hands and eyes already busy, so the tool can't depend on a screen. Any tool that does this has to get information to workers without pulling them off the job, and it has to carry what they see back into the systems the business runs on. Voice does both, because workers can talk and listen while they keep working. Zello is the voice-first Frontline OS for exactly that reason. See what a [voice-first Frontline OS](https://get.zello.com/frontline-os?hsLang=en) does for workers at the physical edge, and why the rest of the business depends on it. [Previous Post](https://blog.zello.com/frontline-ai-ella-smart-search?hsLang=en) ##### [Where can AI save time on the frontline?](https://blog.zello.com/frontline-ai-ella-smart-search?hsLang=en) ## Footer Navigation ### Solutions - [Frontline Communication](https://zello.com/product/frontline-communications/) - [Radio Replacement](https://zello.com/solutions/upgrading-to-zello/) - [Dispatch](https://zello.com/product/dispatch-hub/) - [Customer Assistance](https://zello.com/product/kiosk/) - [Management](https://zello.com/product/management/) - [Enterprise](https://zello.com/enterprise/) ### Features - [Pricing](https://zello.com/pricing/) - [Channels](https://zello.com/product/features/channels/) - [Message History](https://zello.com/product/features/history/) - [Emergency Alerts](https://zello.com/product/features/alerts/) - [Transcription](https://zello.com/product/features/transcription/) - [Reliability](https://zello.com/product/features/reliability/) ### Industries - [Aviation](https://zello.com/industries/aviation-push-to-talk/) - [Retail](https://zello.com/industries/retail-push-to-talk/) - [Transportation](https://zello.com/industries/transportation-push-to-talk/) - [Construction](https://zello.com/industries/construction-push-to-talk/) - [Manufacturing](https://zello.com/industries/manufacturing-push-to-talk/) - [Hospitality](https://zello.com/industries/hospitality-push-to-talk/) - [Emergency Response](https://zello.com/industries/emergency-response/) ### Resources - [Customer Stories](https://zello.com/customers/) - [Blog](https://blog.zello.com/) - [Hardware](https://zello.com/hardware/) - [Partner Portal](https://zello.com/partners/) - [Developer Tools](https://zello.com/developers/) - [Support Centers](https://zello.com/support/) - [Trust Center](https://trust.zello.com) - [Zello Advocacy Program](https://zello.com/customers/zap/) - [Downloads](https://zello.com/downloads/) ### Company - [About Us](https://zello.com/about-us/) - [Careers](https://zello.com/careers/) - [Contact Us](https://zello.com/contact/) - [Press](https://zello.com/press/) - [Brand Toolkit](https://zello.com/brand-toolkit/) - [LinkedIn](https://www.linkedin.com/company/zello-inc/) ### Get the App - [Desktop](https://zello.com/downloads/desktop/) - [Android](https://zello.com/downloads/android/) - [iOS](https://zello.com/downloads/ios/) English (US) Choose your languageEnglish (US)Español (LA)Español (ES)Italiano (IT)Deutsch (DE)Français (FR)한국어 (KO) - [Privacy Policy](https://zello.com/legal/privacy-policy/) - [Terms of Service](https://zello.com/legal/terms/) - [Legal](https://zello.com/legal/) - [SOC 2](https://trust.zello.com) - [HIPAA](https://zello.com/enterprise/) - © 2024 Zello Inc. ```json { "@context" : "https://schema.org", "@type" : "BlogPosting", "about" : [ { "@type" : "DefinedTerm", "description" : "Physical work and care and service work, which make up the boundary of what AI can do because both have to be done in person.", "name" : "Physical edge" }, { "@type" : "Thing", "name" : "Frontline workers" }, { "@type" : "Thing", "name" : "Artificial intelligence and work" } ], "author" : { "@type" : "Person", "name" : "Bren Lawrence", "worksFor" : { "@type" : "Organization", "name" : "Zello", "url" : "https://zello.com" } }, "dateModified" : "[YYYY-MM-DD]", "datePublished" : "[YYYY-MM-DD]", "description" : "Physical work and care and service work make up the physical edge, the boundary of what AI can do, because both have to be done in person by someone standing where the work is.", "headline" : "The physical edge: where do workers fit in the age of AI?", "image" : "https://blog.zello.com/hubfs/marine-service-lead-directing-by-voice.jpg", "inLanguage" : "en-US", "mainEntityOfPage" : "https://blog.zello.com/where-do-workers-fit-in-the-age-of-ai", "mentions" : { "@type" : "Thing", "name" : "Frontline OS", "url" : "https://get.zello.com/frontline-os" }, "publisher" : { "@type" : "Organization", "name" : "Zello", "url" : "https://zello.com" }, "url" : "https://blog.zello.com/where-do-workers-fit-in-the-age-of-ai" } ```