2026 is a transition year. That was the consensus from the panel — not a prediction, but an observation from leaders who are living it daily inside AI-native companies. The possibilities have expanded roughly 10x since the original ChatGPT launch. The challenge is not accessing the capability. It is figuring out how to absorb it without losing the people, culture, and operational cohesion that make the organisation function.
At Running Remote 2026, CEOs from SmartSites, Glide, and MeBeBot sat down with moderator Liam Martin, Co-Founder of Time Doctor, for one of the most candid conversations about what AI transformation actually looks like from inside growing distributed companies.
The compounding advantage is real
The opening observation set the tone: the gap between companies that are genuinely integrating AI and those that have ChatGPT subscriptions is already widening significantly. The teams investing seriously — in terms of infrastructure, culture, and learning — are pulling ahead. That advantage compounds. What is a gap today becomes an insurmountable distance in 18 months.
SmartSites grew 30% in output with only 5% headcount increase. At Glide, the best engineers no longer work on the product directly. They build the software factory — the AI-driven development systems with feedback loops — that enables the rest of the team to deliver. The concept of the dark factory is being actively built: specifications written by humans, everything else executed by AI.
Every role has changed
The panel was clear: every single position has changed at AI-native companies. Software developers, customer service agents, and finance and accounting roles are experiencing the most dramatic transformations. SmartSites analysts can now compare performance across 40 client accounts in minutes, delivering insights that previously required days of work. The data always existed. The capacity to use it did not.
Client-facing roles like sales and project management have seen less radical change, because human meeting time remains roughly constant. But AI is already improving preparation, follow-up, and cross-client pattern recognition. The volume of high-value work a single person can produce has expanded substantially.
The hardest part is unlearning
Not everyone has made the transition smoothly. Engineers at Glide quit during rapid AI transformation — not because the product changed, but because their personal identity and pride in code aesthetics was directly challenged as those skills became automated. Engineers who defined themselves by the craftsmanship of their output found the transition existentially uncomfortable, even when the business case was obvious.
The key challenge is unlearning current processes and behaviours. Organisations need cultures where continuous learning is expected and natural, where employees can abandon approaches they spent years mastering, and where the question is not “are we changing?” but “are we changing fast enough and together?”
Glide rebuilt their product twice in six months after spending five years on the original version. Each new generation of foundation models requires recomputing — deleting prompts, rethinking tools, rebuilding feedback systems. This is not a one-time migration. It is an ongoing operating condition.
What AI-native hiring looks like
The panel shared distinctive, practical perspectives on hiring criteria. SmartSites uses the question “How many gallons of gas does the United States use per year?” to assess critical thinking rather than hard skills. The answer does not matter. The reasoning process does.
Glide looks for fearlessness and hunger — willingness to try new things combined with genuine critical thinking. They actively caution against candidates who would outsource 100% of their thinking to AI, because AI tools still make frequent mistakes and judgment cannot be delegated.
MeBeBot asks candidates what they call their LLM. The panellists’ assistants have names: Gino, Feynman, CLU. Someone who has spent real time with AI tools has typically developed a relationship with them. That hands-on familiarity, particularly across personal projects, signals the natural curiosity that AI-native culture requires.
The principle everyone agreed on: job descriptions have not caught up. Companies are building future org charts without updating current roles to reflect the AI orchestration, agent observation, and audit trail management capabilities that these positions now require. That mismatch creates a hiring problem before the interview even starts.
Practical recommendations
The panel’s Monday morning list was direct. First: adopt Claude or a comparable AI and systematically connect it to your systems — documents, HR platforms, data sources. Each integration creates compounding benefits. Second: establish an AI usage policy that addresses shadow AI, provides clear guidelines, and inventories solutions to avoid redundant spending. Third: announce an all-hands AI demo session. The innovations happening quietly inside your organisation are often more sophisticated than leadership realises. Peer-to-peer learning is far more effective than top-down mandates.
On governance: maintain human in the loop for all client deliverables. SmartSites requires developers to review all AI-generated code changes before committing to GitHub. That standard of accountability does not slow things down — it creates the trust that allows the pace of change to accelerate.
Managing the pace
Change fatigue is real. The moderator described it as a supersonic tsunami — every week feeling like a month. MeBeBot addresses it through humour and storytelling. Glide runs customer visits when teams get overwhelmed by social media’s relentless “you’re behind” messaging. Meeting grateful customers who are getting real value from the work regrounds people in what matters.
What the panel kept returning to: everyone is navigating this simultaneously. No company has figured it out. The organisations doing best are the ones that have accepted uncertainty, built strong enough cultures to absorb it, and kept people genuinely connected to each other while the ground shifts underfoot.
