At Web Summit Vancouver, our User Experience, Software Development, and Creative teams discovered that while everyone’s chasing the latest technology, the real differentiator still comes from thoughtful implementation, quality user experiences, and solving genuine problems for people.
Jacqueline Antalik
Director of UX and Service Design
Shaping an ethical and responsible future for AI—we all have a role
Many of us are experimenting with AI in our daily work and lives. While it’s easy to feel like we’re falling behind—especially when everyone’s sharing how they’re using AI—a key insight I took from Web Summit is that despite all the excitement, AI is still in its infancy.
Several presenters talked about AI automating repetitive tasks or augmenting human capabilities, but noted we’re not quite there yet for handing over the reins. While many look to AI for efficiency gains, a 100x increase in output doesn’t guarantee good outcomes—per AI slop.
A few current shortfalls of AI include:
- Generative AI: In “Designing the future: The new visual workflow,” the speakers shared that the current image quality output doesn’t satisfy sophisticated users, not even beginner users. Many tools have issues incorporating text with images. AI technology hasn’t reached the point where you can simply upload a few images, such as a branded poster, and receive an output that aligns with your brand’s standards. Interestingly, companies like Pixar are collaborating directly with AI firms to train brand-specific models. While this approach is beneficial for organizations with the budgets to support such initiatives, it remains inaccessible to most small-and medium-sized businesses.
- AI agents: In “The agentic era,” the speakers noted that today, many people feel uneasy about giving control to AI agents, even for simple tasks like booking a flight. While AI agents aren’t advanced enough for complex, nuanced tasks (which are often less binary), we are getting close. In the meantime, as we look for scenarios to leverage AI agents, it is essential to consider two key questions to determine if a particular use case is suitable for an AI-enabled solution: How can we ensure AI accountability, and what are the liability implications of using an AI agent?
Despite these shortfalls, now is the time for us to be active participants, not passive observers. It’s essential for all of us to engage actively in shaping the future of AI. AI is here to stay and will significantly impact our lives. We all have a role in ensuring AI is applied ethically and responsibly in our workplaces and society.
James Byun
Art Director
The differentiators haven’t changed: quality user experiences and solving real problems
Web Summit made one thing clear: as AI grows, we need to think more strategically. It’s easier than ever to build with AI—there are tons of new AI startups and products, and anyone with an idea can jump in, even without a tech background.
For digital services, this means AI will soon be the baseline that users expect, not a special feature. Our advantage stays the same: building quality user experiences and solving real problems for people.
But more AI also means more security risks. We need to keep focusing on security, privacy, and ethics when we build AI tools—especially since people need to trust the services they use.
The biggest change? How people look for information. They don’t just search on Google anymore. They ask ChatGPT, check social media, or browse forums for answers (Search Everywhere Optimization). This means we need to rethink outdated past best practices. People who grew up with smartphones want experiences that feel authentic, transparent, and easy to research on their own.
It confirms that our job isn’t to chase every AI trend—it’s to make sure digital services meet people where they already are, with the same quality they expect from their everyday digital experiences.
Steve Ly
Director of Software Development
Agentic AI is coming fast. Are we ready?
One main takeaway from Web Summit 2025 in Vancouver: Agentic AI is coming fast and will reshape how we work. Agentic AI refers to systems that can autonomously make decisions and execute tasks. Unlike regular AI tools, agentic AI systems can make decisions and complete tasks on their own. Think of them as digital teammates, not just software—but with increased risk for enterprises.
Many speakers emphasized that integrating AI agents (or any AI tool) should be treated like new or junior employees. You need clear policies, oversight, and accountability structures. You wouldn’t hire someone without training or supervision—AI should be treated with the same care.
Other reflections:
- The lure of agentic AI is strong, but experts advised to go slow. Start with deterministic workflows and only introduce autonomy where it adds value.
- As AI gains more decision-making powers, organizations need policies that ensure ethical use, security, and alignment to core values.
- Based on the number of startups and established players at Web Summit, tooling is clearly outpacing adoption. The question isn’t what’s possible anymore—it’s how to use these tools responsibly and effectively.
While AI dominated conversations, quantum computing also got some of the spotlight. It’s exciting to see early momentum building. Quantum and AI will increasingly influence each other as both technologies mature.
Wil Arndt
Creative Director
A creative strategist’s take on practical AI implementation
One of the interesting takeaways I synthesized from several speakers across multiple sessions at Web Summit Vancouver was how AI agent development is becoming less about the underlying models (which are quickly commoditizing, according to Douwe Kiela) and more about thoughtful integration and governance.
As a creative strategist, I was struck by the practical framework that emerged: start with simple “copilot” agents that assist rather than replace human decision-making, then gradually build toward more autonomous systems only after proving value and establishing proper oversight (May Habib defines this in four levels of agentic AI).
The real competitive advantage isn’t in having the fanciest AI—it’s in how well you connect it to your specific data and workflows, while maintaining transparency about what the agent is actually doing and why (David Shim recommends contextual citations, while Alex Gallego recommends leaning on the concept of “dead letter queues” for agentic development).
Most importantly, there always needs to be a human ultimately accountable for the outcomes, even as the technology becomes more sophisticated. For those of us in creative and strategic roles, this suggests AI agents will be most valuable as collaborative partners that enhance our capabilities rather than black boxes that make decisions for us (Though Hootsuite’s Irina Novoselsky claims agents are already in her company’s org chart).
Ultimately, it’s not just plug and play (yet). There’s still a strong need for systems thinking and a clear intentionality around the things we want our agentic partners to help with, and how we expect them to do that.
In summary
Our Web Summit takeaways reinforce what we’ve always believed: the best digital services start with understanding real user needs and solving genuine problems.