Colin Fitzpatrick

I write about AI, Web3 and Emerging Technology. 10k Followers. Co-Founder & CEO. Ex Salesforce, Oracle, Hubspot & Dell. Host of the Web3 Leaders Podcast. Featured on CNBC, Bloomberg & Fintech.tv #ai #blockchain

Founder, entrepreneur, storyteller, speaker and Podcaster. After two decades in senior roles at Oracle, Salesforce and HubSpot, Colin moved to Dubai to Problems of the world of startups, helping companies on strategy, storytelling, go-to-market, and zero-to-one growth. A technologist to the core, Colin pivoted into Web3 in 2020, launching a metaverse and NFT token 20 a 300m market cap on three top exchanges, working with global icons such as Snoop Dogg and Alicia Keys. He has since founded several AI ventures—including a conversational 3D avatar platform for always-on customer support - and has raised more than US $10 million across his projects.With 25,000 followers on LinkedIn, Colin is recognised as one of the leading voices in artificial intelligence. He has been featured on CNBC, Bloomberg and FinTech TV, and he has extensive experience as a keynote speaker and panellist on stages at major fintech, blockchain and AI conferences across Europe, the Middle East, Asia and North America.Beyond building companies, Colin advises early-stage founders on fundraising and go-to-market strategy, and is also a podcaster exploring the crossroads of AI, Web3 and finance.His trademark energy, knack for turning complex ideas into plain English, and compelling storytelling make him a sought-after speaker and mentor.

Artificial Intelligence Consultant - UAE Partner

An OpenAI engineer says he doesn't write code anymore. The week after Anthropic used Claude Code to write 100% of Claude Cowork in a week.

The CEO of developer platform Cursor said they let AI run completely on its own for one entire week, utilizing several thousand agents, to build a full web browser, and it did it. 

It wrote 3 million likes of code on its own. This would take an entire development team several years in comparison.

However, he said 'it kind of works' - bit renders pages super fast, but it's certainly not production ready.

But in a week?

A year ago it was barely writing a few hundred lines reliably. 

The writing is on the wall. Things are about to change.

Now, please do not be mistaken. This does not mean devs are no longer needed, or anyone can build serious apps.

When the calculator was invented it did not make everyone a mathematician.

You need to know how to instruct these tools and that takes decades of experience.

But several guys I follow like Mccay Wrigley says he no longer reviews code in Claude Code he just clicks approve. Several friends I know have said exactly the same thing. 

It's that good. 

From my own experience, 6 months ago I was tearing my hair out with Lovable and Emergent trying to build something, but every new thing I did would break something else. 

Today, it doesn't really make major mistakes at all, and does eveything I need. 

Next we'll see the same thing happen with tools that can do non coding work too. 

Strap in.

Founder. Entrepreneur. I write about AI and Emerging Technology. Speaker, Presenter & Podcast Host - Featured on CNBC & Bloomberg #ai #agents #bitcoin

I stayed up until 2 a.m. last night vibe coding on Tuberizer and honestly I’m still kind of blown away by how much you can ship in such a short period of time.

In one day I added a stack brilliant new features. Some are just small but important UI and UX improvements - toggles for summary length, text size, timestamps, but I even added an LLM style “chat with the video” chatbot where you can query the transcription. 

You can also download transcripts and summaries, share directly to socials, and there’s a really slick feature where if you take any YouTube URL and just replace the word 'YouTube' with "Tuberizer" - it brings you directly to the summary share page. Magic!

I then set up a full paid premium tier with Stripe - monthly and annual plans - and gated features cleanly between free users, registered users, and premium, with multiple upgrade prompts included.

I then created a full admin dashboard tracking traffic, growth, conversion, and the content, while adding some scans for security, summary errors and more.. 

The hardest part ended up being something small but critical. When sharing links, the video thumbnail and metadata wouldn’t show. Lovable doesn’t support this natively, so after hours of troubleshooting I eventually solved it using Cloudflare Workers and routes, guided step by step by Claude. It tool many hours but that moment when it finally worked was ridiculously satisfying.

A year ago, this probably would have taken a developer days or a full week. I did it just by explaining what I wanted, sometimes asking for recommendations, and sometimes it gives me ideas I hadn’t even thought of.

Now let's be clear here. Every time I write about Vibe coding, I get salty developers mocking me in the comments saying I don't know what I'm talking about and coding is not software engineering. 

To be clear, this doesn’t mean experienced developers aren’t needed anymore. Quite the opposite actually. 

What’s really happened is the barrier to building MVPs has collapsed. That will lead to 10x or 100x more people building apps, and an even bigger need for seasoned developers to make things robust, secure, scalable, and fit for purpose. Vibe coding won’t replace that.

There’s also a lot of noise about SaaS companies being in trouble because businesses will just build their own tools instead of paying for Salesforce or SAP. That misses the point. 

Enterprises don’t pay millions for features. They pay to outsource extreme complexity, compliance, and liability across jurisdictions. That isn’t going away.

The wall to building your own app has now fallen flat - at least at the MVP level.

Marketing and distribution was always the hard part. Now we just get there much faster.

And in an attention economy where everyone is fighting for eyeballs, that might be the real bottleneck going forward.

It’s the weekend, are you planning to watch some YouTube? Check out Tuberizer.com - it gives you clear channel summaries so you can decide what’s actually worth your time.

We’re all drowning in new content. It’s impossible to keep up. And when you do commit an hour to a long-form video or podcast, you want to know it’s going to be worth it.

That’s exactly what I built Tuberizer to solve.

What you can do with it:

- Summarise any YouTube video or entire channel
- Toggle between short and long summaries
- Save channels and videos to your favourites
- Log in with Google to get a custom feed
- View videos by latest release or grouped by your own categories

You can also create shareable links that show a full summary before someone watches. 

How often do you send someone a one-hour video and they never watch it? Usually because they don’t know what’s in it or if it’s worth the time. A summary fixes that instantly. It’s also a great way to discover new channels.

What’s coming next:

- An overall summary of what a creator has covered across their last ~10 videos
- Category-level summaries, especially useful for AI and tech

I’ve already had requests for crypto and finance, and I can add more categories easily. No matter what, you can always summarise any video or channel yourself.

If there are features or categories you’d like, let me know and I’ll try to make them happen.

And thanks so much to the people who’ve reached out saying they’re using it every day. It genuinely means a lot.

Like the Idea? Please give me a Like 👍 ,  Comment 💬, or better Share ♻️  to help me out :)

taplio

Colin Fitzpatrick's Linkedin Analytics