Is it just me, or are tech events and communities on the rise?
I’ve been thinking about this lately, and I’d genuinely love to hear your experienc...
For further actions, you may consider blocking this person and/or reporting abuse
Great article ! I would said that tech events and meetups are on the rise in Toronto or Ottawa. Ottawa has TechTuesday. I spoke to the president of the chapter of Ottawa TechThursday where it is an affiliation of TechThursday in Alberta. I find that it is great place to share ideas and network opportunities because you can meet founders, manager etc.... at the local tech hub.
Thanks! Great to know! I will check if Toronto also has such chapter in the city.
I think that you ask me about Gemini 3 last week. I find that Gemini 3 a bit better than Gemini 2.5 flash for a project that I am done this week. It will depend on the context for the project that you use Gemini 3 flash.
No problem! I find these events really valuable because you get to learn about the latest technologies and meet guest speakers who are founders from across the region and even outside Canada.
Thanks for sharing your feedback on Gemini 3! Good to know. How about the cost per token? Does it vary much?
They have no cost for Gemini 3 flash. It is a free tier. If you want to paid for an upgrade model. They have Gemini 3 PRO. Google came out with an update version of Gemini 3 PRO a few days ago.
Ok great, thanks.
i can feel the changes drastically after i cahnge to remote work. now i not just a software engineer, i now have to do everything from design to deploy and im a one man team. really quite challenging for me because vibe coding doesn't help much and im broke to use any paid AI.
Agree. The "age of AI" is just a simplified monicker hiding more complex changes in society that have been going on for longer already. Take any 2016 post about algorithms and replace "algorithm" with "AI". Chat bots, automization, remote work and globalization make many people spend their work isolated in front of a computer or a factory machine the whole day, while more people actually worked together with other people in earlier times.
Paid AI doesn't live up to its promise. Most services let you test paid plan previews and you can see they're still flawed and prone to errors nevertheless. It's just that you get more computation time and don't rapidly run out of tokens. But if your one man "team" has different team roles, they might all have their own accounts and email adresses and browser profiles for sure? Thus each one of them has another fresh free AI session when their "coworker" ran out of tokens? Just an idea so you can better evaluate what you are missing out or not.
I can relate to the difficulties of faster context switching and the urge to become competent in more domains, not just software engineering. I would say that your competitive advantage is that you can understand software more deeply, being a software engineer by trade, so you can leverage this to build higher quality software and maintain it better over time.
I appreciate that you are adapting and expanding your knowledge. Curious, what is the hardest new skill you are learning? Design, marketing, or other?
I wish there were even more real-life events but I agree there is a shift, maybe finally catching up with pre-pandemic levels without its everyone-back-to-office culture though. Berlin and other big cities in Europe buzz with events, not only in tech, and people crave for authenticity and real-life experience that matters.
Agree! I was living in Berlin until last year and I really enjoyed the events there. And yes it's not only in tech, people are seeking authenticity as you say. A signal for me is a big uptick in manual crafts making workshops which have become very popular in my area.
Interesting perspective, and I appreciate the sincerity behind it.
I'd push back gently on one assumption though—that fast execution creates a validation gap only community can fill. For practitioners building governance and compliance systems for clients, the validation isn't social. It's structural. Does it hold under adversarial conditions? Does it degrade gracefully? Will it still work in two years when the person who built it isn't in the room?
The "build and demo in two hours" culture is exciting, but it optimizes for applause, not durability. My clients don't need me to attend a hackathon—they need a solution that sits on top of their existing systems without creating new dependencies, and that works without my direct involvement going forward.
Where I do agree is on empathy. But I'd frame it differently—empathy isn't a soft skill that makes AI better. It's a design constraint. If your governance layer can't account for how humans actually make decisions under pressure, it fails regardless of how elegant the architecture is.
AI didn't lower the barrier to building. It lowered the barrier to shipping. Those aren't the same thing.
Thanks for providing these insights Narnaiezzsshaa, I appreciate the nuances you raise here.
Yes I agree with your point regarding governance and compliance systems that require structural validation vs social. I am not claiming that these fast-paced build events are suitable for all types of systems, I am simply raising a general trend I am observing that there is an increasing number of such community events in my city. Nor am I saying that these events produce production-ready systems, but at the least they provide early validation signals.
I really like the framing of empathy as a design constraint. But don't you think that building empathy as a soft skill can help subsequently design the system with the right constraints?
Could you also expand on your last point? With coding assistants and agents generating code for you, I would argue that AI lowers the barrier for both building and shipping no?
On empathy: You're treating empathy as a human input that feeds into design—a soft skill developers cultivate and then apply. My position (and EIOC's foundation) is that empathy is a system property, not a developer trait. The distinction matters because soft skills depend on who's in the room. Design constraints persist regardless of who's operating the system. If your governance architecture requires the right person to be empathetic for it to work, it doesn't work. Empathy has to be embedded in the structure, not hoped for in the operator.
On building vs. shipping: AI does lower the barrier to generating code and deploying it. But "shipping" in the sense that matters—delivering something that holds under adversarial conditions, degrades gracefully, survives contact with real users, and functions without the builder in the room—requires governance, validation, and accountability. None of that is accelerated by AI code generation. You can push to production faster, but pushing to production faster without a governance layer just means you fail at scale instead of failing locally.
Got it, makes sense to me now. I fully agree with the failing at scale vs local shift. Thanks for clarifying these points.
Great post Julien! This resonates a lot with what I’ve been noticing too. It almost feels like the faster and more independently we can build with AI, the more we need spaces where that work can be seen, questioned, and understood by other humans. Shipping alone gives output, but meaning seems to form in conversation. I’ve personally found that writing and interacting here has created that same effect in a quieter way. It makes me wonder if communities won’t disappear because of AI, but actually become more important as anchors for identity and perspective in a world where execution itself becomes easy.
Thanks Aryan! I agree, writing here has also the same effect for me.
Exactly, I like the way you phrased it. There does seem to be a signal shift from execution to community building since the former is getting faster.
This resonates deeply. In SF, we're seeing the same surge — more hackathons, more AI meetups, more "build and show" nights than ever before. I think you nailed the paradox: the faster and more independently we can build with AI, the more we crave real human feedback loops. Speed without validation is just shipping faster in the wrong direction. The builders who will win long-term aren't the ones moving fastest solo — they're the ones embedding themselves in communities that challenge and sharpen their thinking. Great read, Julien!
Thanks for sharing that Vic! Nice to hear similar dynamics happening in SF, this is a meaningful signal.
Exactly — and I think it’s especially telling that it’s happening in tech hubs. The people building AI are often the first to realize what it can’t replicate. Curious what forms you’re seeing it take in your context?
I am seeing a lot of hackathons and open build events. I am also seeing companies opening up their office space on some days for external builders to come and connect together. Which I feel is a great thing for innovation and sharing knowledge/ideas.
That's a pattern I've been noticing too. The office-space-sharing thing is especially interesting — it signals companies actually see IRL cross-pollination as a feature worth investing in, not just a perk.
Some of the most useful conversations I've had building in SF happened in exactly those setups: a random hallway chat with someone on a totally different problem, and suddenly you have a new angle on yours. That serendipity is really hard to engineer intentionally, but creating the physical conditions for it is apparently something people are getting better at.
Hackathons serve a different purpose but the energy is similar — you're building with context, surrounded by people who get it. Both feel like the ecosystem self-correcting.
Excellent post!
Yes!! What people are looking for, particularly with AI, is a human-like exchange with human experiences. This is why I started writing companion prompts. The first step in getting AI to be more human is through clear prompting. AI has no idea what we want unless we tell it. I'm excited about future developments in this niche. I have had great success with this using DeepSeek AI.
Thanks Luc!
I couldn't agree more!
Very nice, I am always looking at ways to improve the quality of my prompts. Do you have any tips or resources to share for quality prompting?
I'm actually working on some prompt packs now -- specifically for deeper connection and meaningful conversation. I'd be happy to share a couple. What kind of conversations are you hoping to have? For me, the key has been specificity -- instead of "be my friend", try "tell me something you've noticed about yourself lately." Also, prompts that invite vulnerability tend to go deeper. What's been working for you?
Nice! Sounds interesting. I would love to see a couple of your prompts.
I agree with the specificity part.
I mainly prompt in the context of coding. What has served me well in terms of prompting are
Outside of coding, I would be interested in setting up a personal workout plan for example. What kind of conversation should I engage in this case for example?
Here are some examples of how I would handle setting up a personal workout plan:
"I'd like to create a weekly workout plan. What should I consider first — my goals, available equipment, or current fitness level? Walk me through how to start thinking about this."
"If I have 30 minutes, 3 days a week, and only dumbbells at home, what kind of effective full-body workout could I build? Give me an example routine."
"I'd like my workout routine to also support my mental health. How can I structure exercise to reduce anxiety and improve mood, not just build muscle?"
Ok I see, you are asking for structure and clarifying goals here. Good learnings. I will use these prompts. Thanks for sharing!
This is why we are launching the Cloudinary Creators Community. It's online, but who knows what in-person events we might launch in the future! I'm making a bet that human connections are still valid. community.cloudinary.com
Love it! Thanks for sharing.
This really resonates with me.
I'm building an open-source AI coding tool on my own. Technically, things move forward every day. But without feedback from real people, it sometimes feels like I'm just building for no one.
I got into this because I believe AI can make coding accessible to everyone — not just experienced engineers. That's exactly why I started showing up in communities like this one. Not to sell anything, but to find people who care about the same thing.
The line that stuck with me most was: meaning is rarely created in isolation. You can ship fast, but if no one reacts, challenges, or validates what you've built, you start to lose sight of why you're doing it.
Thanks for writing this, Julien. It put into words something I've been feeling but couldn't quite articulate.
Thanks for the feedback and validation!
The distinction between shipping and meaning is the right one.
There is a specific moment in any building process where optimization for output gives way to needing actual response — not validation exactly, but the feeling that something you made has become part of someone else's thinking. That threshold is underbuilt in almost everything we design. Most systems handle consumption well. The crossing from passive user to active contributor is where things break.
Something I notice in my own context: receiving input isn't the same as participating. You can interact with a system for months without either of you being changed by it. The events you describe create conditions for the other thing — where showing up becomes leaving something behind.
That's the piece AI doesn't accelerate. The speed of building goes up. The speed of mattering to someone doesn't.
Well said Kalpaka! I totally agree.
This hits home. I run 7 AI agents as my business team and the human connection piece is something I think about constantly.
Here's a real example: I added my AI chief of staff (Rocky) to a Telegram group chat with a field employee. The employee sends photos of receipts and checks. Rocky processes them, asks follow-up questions, coordinates with the AI accountant.
The employee has no idea he's talking to an AI.
And I caught Rocky breaking character — using formal English instead of casual local language, asking redundant questions, posting internal notes. The "human" mask kept slipping.
It works operationally. But it raises an uncomfortable question: as AI handles more of the communication layer in businesses, where does the actual human connection live? Right now I'm the only human in a team of 7. My customers talk to Raven (human sales rep). But between me and the operational layer? It's all AI.
The need for human connection isn't just philosophical — it's a real design problem when your team is artificial.
Interesting, thanks for sharing Warhol. I would imagine that traceability is crucial for these setups and feedback loops.
Curious to know what measures you are putting in place to mitigate some of these difficulties you are facing?
This hit home as a cloud engineer who just attended their first local tech meetup.
One thing I'd add to the conversation: yes, AI is lowering the barrier to build — but it's also flooding the space with new tools every single week. For beginners, that's overwhelming. Less opportunity, more paralysis.
Maybe that's another reason these events matter. Not just to connect, but to get clarity from people who've already navigated the chaos.
Do you find that the events you attend help people focus — or do they just add more to the pile?
That's a really good point you raise. The increasing choice of tooling available is can feel overwhelming and people want clarity and reassurance on their choices. I have noticed that a lot of conversations resolve around the tooling setup (which models, frameworks, MCP/skills etc.).
I find events are a way to get clarity but also to expand your horizons and get inspired by new ideas and opportunities.
Noticed the same thing — the faster I can prototype solo, the more I need someone to poke holes in it. Speed without feedback is just building faster in the wrong direction.
Frequent human feedback is very valuable indeed. Do you have any strategies in place to gather human feedback?
Imagen brilliant hipperrealistic luxury oil style paintings of Rembrandt The Night Watch with deeply realistic colors in brutally creative design