Stop passive technical monitoring: 5 ideas to set up a tech intelligence system
Like many of you, I do technical monitoring on a regular basis. But I’ve never really taken the time to rethink how I go about it. This article is a reflection on what I’ve started to put in place to improve the way I explore new technical tools and concepts.
Let start with an observation: we tend to look for what’s trendy and follow everything that’s new. Why do we do this? Because of Fomo (Fear Of Missing Out), so as not to look like an idiot to other people, nor miss out on an opportunity. This is the case with LLMs and Generative AI. Most people want to know a lot about it without any strategy about how to explore this new field. What happens is that we follow all the posts of the trendiest people, skim the blogs and articles in the trade press, subscribe to dozens of newsletters.
Finally, we waste time on minor subjects, read a lot of posts in endless news feeds, and don’t get much out of the process.
What I’m proposing is to regain control of your technical watch by adopting an active attitude when faced with novelty. Here are the 5 key points of this approach.
1 Paradigm shift: start with a need
The idea is to focus your research on a problem you’re encountering in a project, or on a day-to-day issue. The first advantage is to filter the information so that you can devote more time to a smaller number of subjects. As the saying goes, there are only 24 hours in the day. We can’t keep up with everything, so we probably should concentrate on what’s of value for us.
What’s more, it allows us to form a more pertinent opinion. Indeed, many technologies are promoted by their creators and presented as being great. This is legitimate, but does little to help us know whether a new tool should be followed up and added to our toolbox. Having a problem we’re trying to solve allows us to assess in a concrete way whether a tool is valuable.
Ideally, each technical innovation should be tested on a real-life use case.
For example, I recently tried to extract data from a film festival agenda in png format. I tested several options and discovered new tools, some of which were very interesting on paper (based on trendy LLMs etc.), but inoperative in my case. So I’ve learned twice: I’ve discovered tools I didn’t know existed, and I’ve understood some of their limitations.
2 Customize your monitoring
As you’ll have gathered, the quality and value of a technology can be appreciated differently depending on how it’s used. That’s why it’s important to be clear about what you need, before you start searching all the blogs you know. If I need to know the state of the art on a computer vision topic for complex “inpainting”, I might look for scientific publications.
If, on the other hand, I’m looking for a solution that scales up and is stable, I’ll look for open source projects with many contributors, for example.
We can ask ourselves a few questions to fine-tune our technological intelligence system on what is important to us:
- 🎯 Goal. What am I trying to achieve? Improve my security system, scale my software, add new features to my solution, etc.
- ⛳️ Field. To do so, what field do I need to explore? Who are the key players and influencers in this field? (NVIDIA? Meta? HuggingFace? …).
- 🖥 Media. What media I am more confortable with? Podcasts (because I can listen to them running), blog posts (which I can read when I am waiting for the dentist), videos (which I watch just before going to bed), scientific paper (beceause I want to deep dive into a subject), etc.
3 Get organized
Now that you are clear with what you want and what you need, you can start to organise your technical intelligence.
Gather the source of information. The following questions should therefore be answered in the light of your own needs to determine which source to use:
- Who to follow on social networks? One can search for trends and most followed people on social media.
- Which site to visit regularly?
- Which blog to visit regurlarly?
- Which newsletter to subscribe to?
As far as I am concerned, I prefer people who seem to be passionate about sharing and do it mostly for the love of tech. There’s often a promotional dimension to the publications, which is quite legitimate, but it’s something to keep in mind if you want to remain critical of what you read.
Some Newsletters are really good and match my expectations. However, I limit the number of newsletters I receive, as they are often very general and require a lot of reading time for a small amount of information that is useful for oneself.
Schedule your work of intelligence. In my experience, the best way to do something regularly is to set aside a dedicated slot in your diary and add a measurable goal. For example, someone could book a Monday morning slot with the aim of reading 1 complex article and testing a new python library.
We should make sure to have enought time for reading/listening, but also for testing. Here are some example of agenda slot that can be scheduled for technical intelligence :
- Lonely reading sessions
- Sharing sessions on specific themes
- Mob testing sessions wih our team
Use tools to retrieve easily information. We often hear about a new technology or come across an article at a time when we can’t go any further. What’s more, some subjects are complex and need to be revisited several times to become fully understood. That’s why I think it’s important to be able to easily find the information we’ve come across throughout our technical watch. To do this, it may be useful to set up a few tools:
- A glossary to help you memorize new concepts
- Lists of articles, podcasts you wan to read or listen to, so that when you have free/waiting time you don’t loose it
- Diagrams to easily represent complex ideas and combine them with others
4 Talking to people
This is the way I learn the most. Rather than thinking that the person you’re talking to will think you’re rubbish, we have everything to gain by asking questions about the best practices, discoveries and tools of our colleagues, acquaintances and so on. It enriches your conversation, you have the opportunity to dig deeper into a subject by asking questions, and you get live feedback that saves you an incredible amount of time.
Beyond this aspect, it’s an excellent opportunity to forge links, share your passion and, why not, make new friends.
5 Keep an open mind
To avoid going round in circles, I think it’s important to keep an open mind.
To do so, we can try to regularly seek out new sources and sort through the old ones to evolve our system. For example, we can rely on other people and explore the references they use.
In addition, I really like to add articles or podcasts that are more “meta” and talk about personal organization, learning, communication and social phenomena, so that I can continually work on my own practices and try to improve them by exploring new fields.
That’s all there is to it! If you also have ideas for improving your technical watch, I’d be happy to hear from you.
