AI Productivity Gains

In partnership with

Hey 👋

I did a demo working with Operator and Deep Research of OpenAI and Google, comparing them.

Also, while there are hefty productivity gains already possible, v2 and further versions of e.g. Operator will give us insane productivity gains. Because open-source is pushing hard.

[Demo] OpenAI’s Operator can perform multiple tasks in parallel

To run Operator you need to have an OpenAI Pro account (200$/m). If you are from the EU, like me as I am from Munich, Germany, you even need to use a VPN to have access to this tool.

Operator, sort of like your AI EA, is in its first version. And not yet effective at all times. However, I am learning to work with it in the most productive way possible.

In this video I demo how to outsource multiple tasks to an Operator fleet. I hope it helps.

The next version must be significantly better, as there are already open-source versions for close to 0€.

[Demo] Deep Research: Google vs. OpenAI

Both OAI and Google have released a Deep Research feature within their AI tool, ChatGPT and Gemini respectively.

As I have access to both (oh boy, all of these subscriptions .. 🥲), I thought I will do a direct comparison of A) how well their reasearch capabilities are and B) how well they can reason over the findings.

I hope you enjoy it. (Who do you think has the better product?)

(It is OpenAI’s version of Deep Research. 😉 )

Bolt(.new) Updates

If you want to build (AI/ tech) products (which everyone can do these days), I consistently see incredible productivity gains when combining Bolt.new with Cursor’s AI Agent. Bolt(.new) for the coarse-grained things you want to code like UI/ UX/ DB etc., and Cursor for the fine-grained adaptation of the product.

Pro tip: When you are stuck with hard problems you can’t solve in cursor, use OpenAI’s o1 in Pro Mode. With this, you are, hands down, unstoppable.

These tools update their tools btw. very frequently. Literally, every week one of them pushes a minor or major feature.

Especially, Bolt has had a lot of improvement. So, I thought I mention my main ones:

1. Integration of he DeepSeek R1 reasoning model

2. Terminal error detection integration (catches like 35% of all errors)

3. Smart imports (most of the errors come from dependency issues like this)

4. Crazy one: makes it faster, saves tokens, less errors → leverage external assets

5. Awesome tools are being built (one of many examples)

If you stay on top of these three apps/ tools, you are going to be unstoppable.

Awesome time to be alive.

I hope you enjoyed it.

Happy weekend!
Martin 🙇

• I recommend:

  • Beehiiv if you write newsletters.

  • Superhuman if you write a lot of emails.

  • Cursor if you code a lot.

  • Bolt.new for full-stack development.

  • My book - Generative AI: Navigating the Course to AGI.

  • Follow me on X.com.

• AI for your org: We build custom AI solutions for half the market price and time (building with AI Agents). Contact us to know more.

You might like our last episodes: