If, like me, you've made the effort to measure your dependence on US tech, you've probably realized that this dependence is very strong.

Like, really strong.

And if you haven't done it yet, now is definitely the time.

Before we begin, I should clarify that I don't usually engage in knee-jerk anti-Americanism. I grew up under American soft power, nourished by Yankee music, TV series, and movies. I listened to Nirvana and was cradled by Star Wars and Quantum Leap.

But today, for the first time in its history, the United States could well be sliding to the dark side. And our dependency, my dependency, on US tech could be a real problem.

So, I'm going to try to take inventory, measure this dependence, and see how I could reduce it.

And for that, we're going to play a game.

Seeking Balance

We're going to play a game because, I must confess, if it feels like a game, you can get me to do anything, sports, cooking, finance, anything.

A game allows me to have an objective, to break down this objective into side quests, and ultimately to motivate myself to move forward in small steps.

And to make a good game, we need a score that reflects my dependence and that I could compare with others. We could even create a leaderboard among ourselves :)

But first, what exactly is this dependence?

When I talk about dependence, if you take the exercise of listing the applications and services you use daily, chances are you use a lot of US services.

Between phones, home delivery services, ride-sharing or streaming services, it's quite illusory to aim for zero dependence.

And, as I mentioned in a previous article, it's not necessarily desirable; interdependence also has its benefits. BUT we need to find a balance point.

This balance is useful in a negotiation to say, "you can't hassle me about A, otherwise I'll take away B."

In short, we want a score that reflects this balance.

So perhaps we could first simply try to count the number of applications?

First Attempt: Counting the Number of Applications

Let's test with this list:

Service	Country
Gmail, Gdrive, Gdocs	US
Booking	NL
FreeNow	DE
Duolingo	US
Twitter	US
Youtube	US
NordVPN	NL
Telegram	RU

Summing it up, we get a score of 6/12 (50% dependence).

But we immediately see that something's off. Not all these services are equal. Can we really compare Duolingo with Amazon for example, or Twitter?

We quickly get the intuition that something's missing. Naively, the balance I'm trying to achieve should serve two objectives:

• Finance more tech in Europe (or elsewhere by default)
• Limit the data provided to tech giants

It's not the number of services I use that's important, but I must take into account the economic impact, what I pay, and the impact on data protection, both mine and that of my close contacts.

Second Attempt: A Priority Score

To materialize these two criteria above, I'll use a scoring system.

• 5: paid AND with a risk regarding the use of my data
• 4: paid
• 3: free, but with a high risk regarding the use of my data
• 2: free, but with a moderate risk regarding the use of my data

For scores 3 and 4, I qualify the risk as high when the company collecting the data transfers it to other companies in the same group or to partners, and when data from my contacts is also involved.

The risk is more moderate if the collected data is only related to the service I use, for example my scores on Duolingo.

This is better, but again, this score has flaws. Free doesn't mean harmless. And paid doesn't mean harmful.

For example, Duolingo (e-learning) is less dangerous in terms of tech dependence than Twitter. Twitter, which can be used to manipulate elections on a large scale.

I realize there's a third objective:

• Finance more tech in Europe (or elsewhere by default)
• Limit the data provided to tech giants
• Limit harmful effects (large-scale manipulation of opinions, for example)

Let's start over and this time, we'll introduce all these criteria.

Third Attempt: Taking Influence into Account

Let's have some fun, let's create a formula to calculate our score.

Our score is composed of:

• FI (Financial Impact), weighted by their importance from 0 to 2

Value	Description
0	free
1	optional service (I could do without it in any case)
2	essential service that should have a European alternative

• SI (Sensitivity Impact) from 0 to 2

Value	Description
0	no data provided
1	little data provided and only necessary for the service (streaming, for example)
2	lots of sensitive data provided (clues about my private life that allow targeting)

• II (Influence Impact) from 0 to 3

Value	Description
0	not applicable
1	minor influence, via non-targeted ads for example
2	possible major influence
3	demonstrated major influence and regularly used

Influence is important. So, I'm going to weight upward the capacity for influence by multiplying it by 1.5, which I choose arbitrarily. I'll also add an additional factor equal to the influence score x the financial score, because if I choose to voluntarily pay for an essential and potentially harmful service, this should be reflected in my score.

Let's take a few examples:

• If you use Facebook and entrust it with your entire life, your photos, your date of birth, etc... You provide a lot of sensitive information that can be used to target you or your loved ones. That's an SI score of 2.

Facebook has already been regularly involved in opinion manipulation scandals like Cambridge Analytica or Doppelganger, so that's an II score of 3.

Its score is therefore: $0 + 2 + 3 \times 1.5 = 6.5$

• Duolingo, granted it's paid, but I can do without it, the data is not very sensitive, and only ads can be used to disseminate targeted information.

Its score is therefore: $1 + 1 + 1 \times 1.5 = 4.5$

• Twitter is a perfect example of dangerous services, already involved in large-scale manipulation cases. But, if on top of that, I pay for the service (the blue checkmark), I finance this manipulation.

Its score (if paying for a subscription) is: $2 + 2 + 3 \times 1.5 + 3 \times 2 = 14.5$

Now, this is starting to look like something. Let's see what it gives on this list of software:

Software	Score
Gmail, Gdrive, Gdocs	11
Amazon	6.5
Uber	6.5
Duolingo	4.5
Netflix	7
Crunchy Roll	4.5
Linkedin	5
Instagram	5
Facebook	6.5
Bluesky	4
Youtube	4
Chrome	5
Whatsapp	3.5
1Password	3

To calculate our score, which we'll call TDI (Tech Dependency Index), we'll normalize it to give a score out of 100.

The dependency score becomes the sum of US dependencies in proportion to the total:

• $TDI = \frac{\sum_{i \in \text{US}} S_i \times 100}{\sum_{i \in \text{All}} S_i}$

My Dependency Score

If I do this calculation seriously, I get 72% dependence based on the applications I use.

My personal balance is clearly imbalanced.

The highest scores are for Gmail (which I pay for because I have 300GB of data on Drive), and Claude.

But now, I can play.

I know where to start, which quests are worth it (scores above 4 in general) or the easy quests (few points, but little effort).

I've started migrating from Gmail to Proton Mail (which comes with Proton Drive, Photos, Pass, and Proton VPN, all products being open source), I use more local alternatives to Amazon (Backmarket, local shops) and I will now favor FreeNow over Uber. I also plan to stop using Cloudflare to replace it with Bunny.net. I'm going to motivate myself to completely abandon Facebook, which only serves me as the boomer that I am to keep in touch with other boomers.

And the TDI (Tech Dependency Index) can allow me to anticipate the impact this will have. These few modifications will allow me to drop to 55%. And above all, I'm going to continue playing.

And if I've encouraged you to play too, this game can really start to get interesting.

If you want to participate, I invite you to download the file I'm making available (on Proton Drive of course). You'll find some typical scorings for certain applications and then, it's your turn to make the next move.

So, what's your score?

PS: if anyone wants to make a static app out of this, feel free and let me know so I can update it here.