What a junior is for, now that AI does the tasks.
Watch the session
The setup
The entry-level job is the first thing an agency cuts when the year gets tight and the last thing it rebuilds when the year recovers. That has been true through three downturns. What is new is that the work entry-level people used to do — the reporting, the deck assembly, the first-draft pass — is now the work most easily handed to a model. Which leaves a question nobody in an independent shop has a clean answer to: if the tasks that used to train juniors are gone, what exactly are you hiring a junior to do?
This session was built to answer that from both directions at once. On stage: two agency principals, a founder who runs a company for people leaving college, a copywriter who arrived in advertising at 30 from restaurant management, and a currently enrolled university student who spent the past year as an agency intern. Five people at five different distances from the entry-level job.
The 40 minutes produced one hard financial argument, one direct disagreement about whether junior talent needs protecting at all, and a rule for AI at entry level that is specific enough to write into an employee handbook this week.
It also produced a quiet irony that the panel did not name, and this chapter will.
Who was on stage, and why they're worth listening to
Part One — The interview stopped being a one-way interrogation
Katie framed the opening problem plainly: the hiring conversation has historically run in one direction, agency to candidate. Her question was how early-career talent flips it, and vets an employer well enough to avoid walking into a burnout factory.
Kelsey answered first with a diagnostic that is about the candidate's own habits rather than the employer's benefits page. Her claim is that burnout has a leading indicator, and it shows up before exhaustion does.
"We cannot stop creating for ourselves. That, I think, is the number one cue to burnout — when you're creating for others so many times that you're no longer creating for yourself, and losing that creative spark."
That reframes the interview question. Instead of asking about balance in the abstract, you ask whether the job leaves room for the thing that keeps you employable. Kelsey's practical version is blunt enough to use verbatim: ask about learning and development funds, and ask "what is the nights and weekends expectation?" She is candid that fire drills are permanent in this business. Her test is whether anyone above you absorbs them: "I'm very lucky to have a team that is very protective of my time, and I don't even have to fight for it."
Jahleane's method goes around the interview entirely. Her company exists because of a gap she describes in one line that landed harder than anything else in the first 10 minutes.
"When you enter college, you have orientation. But when you leave, you have nothing."
Her fix for vetting an employer is a sales technique applied to job hunting: do not source your information from the people whose job is to sell you the job.
"Ask the people who work there, not necessarily people who are in the application process. You can go on LinkedIn and ask people, 'Hey, do you have fifteen minutes? I would love to talk about your experience being an X at this company.'"
The question she suggests asking them is the payoff, because it tests Kelsey's diagnostic from the inside: what's the last creative project you were able to work on just for yourself? An employee who cannot answer that has told you about the culture without criticising it.
Rachel, fresh from 50 interviews, made the case that the most commonly asked question is the least useful one.
"Balance is a hard one to answer, because I don't know that my definition of what balance is is the same as what your definition of balance is. Burnout comes when you're doing a lot of work without meaning or purpose or connection. So asking questions about the work itself — what kind of work, with who, what kind of support systems do I have, what is the expectation for collaboration — can give you a lot of insight into how the agency is thinking about that."
Katie then supplied the question that works in the other direction, for a candidate trying to find out whether an agency listens: ask what the last piece of feedback from its own talent was, and what the agency did in response. Her assessment of its diagnostic power is that it "says a lot really, really quickly." An agency that cannot name a piece of feedback has not asked for any.
Gilliane Russell is doing this vetting while her feed tells her the job she is applying for has been abolished. Her read on the volume of posts declaring the entry-level role dead is narrower, and more useful, than the panic those posts generate.
"The right place will take a chance on an entry-level person."
She is treating the discourse as a sorting instrument rather than a forecast. A shop that has decided the entry-level job is over has told a student everything she needs to know about it, and she can spend her applications somewhere else.
The session's sharpest correction came when Katie asked what agencies lose by defining junior talent as fresh-out-of-school.
Kelsey's answer starts with a joke and then produces the best argument in the session for hiring career switchers.
"You're leaving a huge blind spot for people who have a treasure trove of experience."
Her evidence is her own decade in restaurants. The claim is not that hospitality is character-building. It is that a brief is a problem about people, and she has more raw people-data to draw on than a colleague who has only ever been in advertising.
"The amount of insights you can pull from people in a place like that — you are seeing people from every walk of life. When I get a brief and we're trying to figure out how we're going to sell a product to somebody, I have so much experience I can just pull from."
She reaches for a specific memory: a customer who found glass in their food, and how she fixed it. Her point is that brand crisis management and table-side recovery are the same skill executed at different scale. "If your brand missteps, how do we figure out how to get back on the right course?"
Then she describes what happened to her as a 30-year-old junior, and this is the passage agency leaders should sit with.
"Even coming in as a junior in my thirties, people really did kind of coddle me. They either didn't want to say the thing they really wanted to say, or they didn't respect the ideas that I was bringing to the table. It limited me in a lot of ways."
Both failure modes are condescension. One withholds the feedback that would make her better. The other discounts the contribution she has already made. Neither is malicious, and both are common, and an agency that hires career switchers without fixing them will keep losing them.
Part Two — The financial case for entry-level talent
The second half turned to money, and Rachel made the argument the session will be remembered for. It is not a moral argument, and she is explicit that it is not.
"Entry-level staff is the biggest unlock you can have as an agency. Not just from a moral perspective, but actually financially. If you can get your entry-level talent to go from this level to this level faster, that's your biggest upside on your payroll."
Read that as an operating principle rather than a sentiment. The margin on a senior hire is largely fixed at the point of hire. The margin on a junior is a function of how fast you can raise their capability, and that rate is something the agency controls. Investment in junior development is the only line in a people business where the return compounds.
Her second argument is about product, not payroll. Corner Table Creative is three years old, and she is direct about what clients are buying when they leave a larger shop.
"By the time they're coming to us, they're looking for something different. So we have to stay ahead of that something different. And entry-level talent is the freshest perspective we have. That is a really powerful thing."
She had the receipt. In a pitch the week before, a client volunteered that they could tell which part of the work came from the youngest people in the building.
"They said, 'You know what? We can tell that your Gen Z team was working on this, because there are insights here for their Gen Z audience that we had never thought of before.'"
The operational detail matters more than the compliment. Those team members were in the room to hear it. Rachel's read: "I don't think there is anything more motivating to that group of people than getting those opportunities." The lesson is cheap to copy. Put the juniors who did the work in the room where the client reacts to it.
Katie asked how you balance daily client demands against the long, slow work of nurturing junior talent. Rachel rejected the premise of the question, and in doing so rejected the premise of the session's own title.
"I don't think the systems necessarily need to be long and slow burning, and I definitely don't think they need protecting. Positioning the work as the enemy is the wrong framing in some ways. Yes, we need to have lives that are fulfilling. But the work is the opportunity. The work is what gets you to where you want to go."
Set that against Kelsey, who spent her closing answer arguing the opposite: that the thing juniors need is precisely protected, subsidised, non-billable exposure. Fresh out of her internship, Highdive sent her to a commercial shoot she had no role on, at the agency's expense.
"I learned years of experience in that moment. Seeing how the creative changed — because if anybody's been on a shoot, it can change in a second. Our talent's not here, there's a snowstorm, they got in a car accident. Seeing how that was handled made me a better creative and a better team player."
And then the line that states the trap most agencies are stuck in.
"It becomes the double-edged sword of, well, I can't send them because they don't have the experience — but they don't have the experience because I'm not sending them."
The disagreement is real and it is not about vocabulary. Rachel's position is that clarity of expectation is the whole obligation: tell people where the bar is, tell them where the path goes, then let the work do the developing. Kelsey's position is that some learning only happens in rooms a junior cannot bill their way into, and someone has to eat that cost deliberately. Both are defensible. A shop that adopts Rachel's framing without Kelsey's spending habit will produce juniors who are told the path exists but never shown it.
Katie's contribution sits between them: structure. She argued for published creative career mapping, so that promotion velocity is a known number rather than a rumour, and for 90-day plans built as shared accountability. Her version has a specific mechanic worth stealing: at 90 days, the employee names three things that would help them in the coming year, and the agency commits to funding one. She cited a retention improvement at an agency that ran the program. (The figure is garbled on the recording, so it isn't printed here.)
Her framing of why the artefacts fail is the part that will be familiar to anyone who has written a handbook: "You read through it and then you never look at it again, and so your workhorse forgets and starts getting resentful. They're thinking that things are not there for them that are."
The AI question came last, from the audience, and produced the most immediately usable policy in the session. It did not come from a founder. It came from Gilliane, describing what her boss at Flock and Rally makes her do.
She was assigned media reporting: interpret the numbers, break them down for the client. Her boss's instruction had a specific sequence.
"First write down your perspective and how you would interpret that. I want you to pop it into ChatGPT or Claude and tell me what AI thinks of that perspective."
The order is the entire policy. The POV — the human's own point of view — comes first, the model second, the model acting as reviewer of a position the person already holds. It preserves the exact skill that entry-level work exists to build, and it takes about 90 seconds longer than the alternative. Her line in the sand is drawn at the point where the sequence reverses: she has heard of interns "so dependent on AI" that the output is "missing a human voice that you're relaying to vendors and clients."
Rachel arrived at the same rule from the top of the org chart, in the most quotable formulation of the session.
"Automate every part of the workflow and process that you can, and do not automate the POV. Teach the POV, protect the POV."
Her follow-up is the coaching question that operationalises it: "Yes, you pulled this information in five seconds. Now what?"
Jahleane's contribution was the limiting principle, and she is unsentimental about the tooling.
"AI should be used as a thought partner, but AI is not that smart. And AI can only pull from what it's already been fed."
Kelsey drew the line by task rather than by principle. She uses models to generate idioms and phrases connected to a brand, work that "AI can spit out for me in five minutes versus me spending three hours of my day." What she wants automated is not creative at all.
"The silliest process is time cards. That's the bane of my existence. It should be taking that away from me so that I have more time to spend on my art."
Four people, four vantage points, one rule: automate the administrative layer and the retrieval layer, and defend the point of view. Nobody on the stage dissented.
The uncomfortable truth
The panel spent 40 minutes agreeing that Gen Z is underestimated. The session's own architecture underestimated the one Gen Z person on it.
Gilliane Russell was asked what shocked her, what she craves, what she feels isolated by, and what students need. Rachel Brandt was asked about payroll economics, growth strategy, and AI policy. That division is not accidental and it is not unusual. It is the standard shape of a panel that includes a junior person: the junior supplies experience, the senior people supply analysis. The student is the evidence and the founders are the argument.
Which is a shame, because Gilliane's strongest material was analysis. Her AI protocol is a better piece of policy than anything a founder proposed, and it was produced almost as an aside in response to an audience question rather than solicited by the panel. She named her own conditions for staying at an agency with more precision than most senior hires manage: not compensation, not title, but whether a small local shop leaves you working alone. "The heart of your work is at being human."
And she said the thing the session was named after, in her own words, without prompting.
"There is this very big blanket of an entitlement kind of thing, and we did go through COVID and all of that, and I think that we are severely underestimated. There are some people that are very unmotivated, and some people that are very motivated. A lot of places do truly underestimate how motivated we can be as students. But we need the human connection to be able to do that."
Jahleane had already supplied the historical correction: "when every new generation comes into the workforce, the overall sentiment is, oh, they're entitled, they don't know how to do X, Y and Z." Rachel added the field evidence, which is the most useful anti-stereotype data point in the transcript. Corner Table Creative had planned to open a physical office in a few years. Its Gen Z employees told Rachel to move it up.
"Our Gen Z team was like, we really want you to open an office faster. They were like, no, no, no, we need to do it now. They come in every day. We have no return-to-office policy. We are now moving offices after a year to a bigger space."
The generation supposedly allergic to the office demanded one, and the founder who started on a desktop computer waiting for creatives to print out Adobe decks was the one being asked to hurry up.
So the uncomfortable truth is not about Gen Z. It is that the industry's condescension is not generational at all. Kelsey was coddled at 30. Gilliane is underestimated at 21. The behavior tracks seniority, not birth year, and it costs the agency the same thing in both cases: the contribution of the person closest to the work and furthest from the assumptions.
There is a second uncomfortable item, and it is about access. Kelsey Williams got to Cannes by dumpster-diving for aluminum cans, stockpiling several thousand of them in her living room, filming the process, and posting it until strangers covered part of her trip. She is a working copywriter at a well-regarded independent. That is what it took. Nobody on stage remarked on it, and the industry should.
What you can use
Every session from Indies @ Cannes 2026 — plus every other Blueprint, the tools, and the surveys — opens up with one sign-in.
