ever1

It's real risk; Under oath before the French Senate, Microsoft France’s Head of Corporate, External & Legal Affairs Antoine Carniaux, said he cannot guarantee European data is safe from U.S. government access, even when stored in Europe. U.S. laws like the Patriot Act and Cloud Act require American tech firms to comply with U.S. authorities, regardless of data location. That means, especially with a current US administration acting against EU interests, that a US based AI solution is not safe.

The dataset is available in Hugging Face: https://huggingface.co/datasets/criteo/CriteoPrivateAd

They are so terrible that they manage to get the 2nd place in arms export... https://www.theguardian.com/world/2023/mar/13/france-challen...

source: trust me bro

Is it due to bringing software dev methodologies into very traditional industries ?

Interesting question but I wonder what would be the risks on the infrastructure in Europe ? IMO, apart from Eastern Europe (and still), the major risk would be on energy cuts. But in this case, that means that you would need to have an autonomous way of producing electricity for running these local ressources. Maybe physical books are more reliable ?

yep, but I guess most people are ok with this, we have a lot of people saying that teachers are priviliged people having too much vacations and not working enough per week. It's a strong common belief. And with current government politics, it looks like we are witnessing the end of a public education in France, the unqualified contractuals teachers hired after a quick interview is more interesting financially than an official.

Why ubuntu would not be recommended today ?

Thanks, the temperatures in the article means nothing for those who don't speak Fahrenheit...

As a french, I would like to see a governmental institution at national or European level which aims to develop open-source software for administrations, police, public hospitals, defence... This would reassure me more about technological sovereignty and the use of taxpayers' money, I am even convinced that it would be much more interesting economically. It could even help to catch up with our technological delay in terms of cloud infrastructure...

Looks like a company with poor tech culture, I am not sure it is specific to France and could be seen in any other places.

Why is that ?

I think I know the place he is talking about: HUBSY café & coworking | Saint-Lazare. Been there several times, you pay by the hour with unlimited drinks and some snacks. It's a small place but it is nice and was life-saving for me.

And there are always 2 hydrogen/oxygen atoms in any molecule

weird, 3BP and especially the sequels were mind blowing for me, among the best SF I have read (And I have read a lot...)

'But when asked about this, Mr Libin has a surprising answer: "What if I paid the person in India the same as I pay someone in the US? Well, why not? Right? Like, why do I care where you live? I just care about how productive you are."'

BS, you can have 2x more skilled engineers for half the price and same quality, logic demands that you do not pay an Indian dev (which will still have greater salary than average in India) the same as US one.

As a French I feel (somehow) ashamed and absolutely not surprised XD.

It is well known that France never had impactful scientists...

What are you talking about ? So far Ariane program seems to be a commercial success. ESA on a scientific point of view too. It is true that Ariane 6 will not be reusable, although cheaper than Ariane 5, just like ALL reliable commercial launchers today (even tomorrow for ULA) outside of space X. It is clear that this is the future. However, these are only the first steps. I am not sure that today space X will be able to be profitable with it. Talking about tax money, note that SLS will not be reusable too...

On the contrary, it would be foolish of the EU to be so dependent on a foreign company, dependent on a foreign law on a subject as strategic as space. It is a crucial choice that, as a EU citizen, I applaud.

source ?

Python: Jupyter, pandas, numpy, scipy, scikit-learn

Numba for custom algorithms.

Dataiku (amazing tool for preprocessing and complex flows)

Amazon RDS (postgress), but thinking about redshift.

Spark

Tableau or plotly/seaborn

Detailed report https://twitter.com/olesovhcom/status/928904373949919232

This mechanism already exists in some way with top journals and peer-reviewing. Their credibility relies on their capability to select the most important and proven papers.

I gave an average, in big companies it depends on which company you are working for, which "Grande ecole" you come from, and what position you have. I would say that the variance is quite high.

Living in Paris, I can feel a strong momentum of entrepreneurship and startups. ex: Recent "Station F" incubator from Xavier Niel is opening in 2018 and will be one of the largest startup campus in the World. Macron is strongly pro-startup. there are a LOT of options for funding (quite crazy when I saw how much we can get with my startup (Banque publique d'investissement, crédit impôt recherche...). It is really easy to start a startup, not much paperwork (surprisingly).

Contrary to 10 years ago, best students coming from "les grandes ecoles" want to be entrepreneur or join a startup. This is a radical change to the mindset from the past when the best opportunity would be in big companies. PhD start to be recognised as they should in private sector and more PhD are leaving academic to private (which I did).

Now, I think that comparing Europe to SV is too premature, (A change in culture is yet to come). But the dynamic is here.

In Paris, salaries in tech startups are high (for what I saw: ~42K junior, ~60 senior (5y)) compared to the cost of living (in France healthcare and education are free), rent is high. So yeah, I am confident for the future of Europe :)

edit: and from what I saw, most startups' products have english language available (interface/doc). Most of them have international expansion in mind.

And Paris ?

At this stage, this book is an introduction to introduction to introduction to machine learning. It has not math/theory background. I'm curious to see the final product. However, I'm worried about the bias to NN. His plots on perf for NN vs traditional algorithms is too general and false depending on the problem.

I was also disappointed by these macbook pro, especially the path Apple is taking toward power users. I searched for a laptop equivalent in quality and performance to the macbook pro and I realised that it is not easy to find at all.(I'm doing Machine Learning so I need compute power.)

If you consider the pro of the mbp:

  -performance : Overall perfs are great on 15", but other laptops are better, especially grahics (but for developpers ?)
   Maybe for cuda in ML. But think of external graphic cards with TB3. Memory big - :(
  -battery : big + for mbp, no competition.
  -screen : big + for retina and ecosystem adapted to it.
  -quality : big +, you can expect a 5y+ lifetime, the quality is amazing compared to the competition.
  -design : big +
  -connectivity: today it's a - for mbp. But in 2y ? Cleary we are switching to all-usbc devices. TB3 gives you such
   bandwidth that even external gpus are possible.
  -storage: big - for mbp, even if it is fast, too small.

So the conclusion is less immediate. If you have in mind that you can sell you mbp with a good price even after 4y, you realise that it is a good investment. For the same quality, screen, battery, perfs the competition have same price or don't even offer an equivalent.

So for myself, as long as macos is poweruser friendly, I will stay with the macbook pro, if it changes, I will move to Linux...with a mac :)

source ?