Self-hosting is a trivial matter if you can roughly map a subset of features of an existing language to your language. Write bootstrap compiler in that language and translator in parallel (as long as you have good AST/LST), run conformance tests on both.

These days with LLM bootstrapping the compiler is a lot like using YACC for the parser. You just point the LLM at your compiler spec, tell it "bootstrap this" and then it does.

Sort of. Although the process may not look exactly like that unless you have a really good spec.

True, I’m assuming you’ve already implemented one compiler. You point the ai to that and instruct the ai machine “bootstrap the language using this reference compiler”.

Sort of. You do not have to implement full compiler before starting building the self-hosted one. With LLM it is possible now to write the spec, and at the same time build both reference compiler and translator, validating your design early. No big upfront effort.

You can, but I still feel like you'd want to solidify the behavior in one compiler before bootstrapping it, otherwise you run into the problem https://news.ycombinator.com/item?id=50342830 was talking about.

The solution is not to feed students from spoon all the time. They are adults and should act as such, it is not a school where it is really important to give every child a chance. Grading homework must be opt-in. Those who want it, get the grade as a feedback, can seek for help then and get the help, those who skip should be assumed prepared for final exam and judged by it. There are millions of people in the world who can use education opportunities to their maximum, and the place should not be reserved to those who feel entitled by whatever privilege they have.

We observe that intervention leads to more students learning. I don't believe that "well they are adults they can fail if they want" is an excuse for a university refusing to take steps that they know would improve pedagogy.

I do agree that AI has made this basically useless. Rather than skipping homeworks or doing a bad job the students just cheat and there is no opportunity to notice and intervene. They get all the way to the screen-free final and fail.

I do not think that controlling homework does improve pedagogy at scale. Homework helps engaged and motivated students who are ready to accept the feedback. Maybe in your place they are majority and effects are visible. As soon as grades for homework become the goal, the Goodhart‘s law starts playing the role in the dynamics of relationships between students and teachers. AI has nothing to do with it, cheating with homework did always exist.

> I do not think that controlling homework does improve pedagogy at scale.

You are welcome to believe this. You are also wrong.

I am curious if you have ever taught at a university.

This particular example is hardly representative. The market they work on is difficult. They may have an ambition to become Tesla of airplanes, but it did not take a long time before traditional manufacturers learned from Tesla’s success and took over. VW builds more electric cars, and number one globally is a Chinese company. With airplanes, becoming an independent manufacturer is a long and risky path. In the end, their main choice is between being acquired by Boeing or Airbus if they are successful, or being outcompeted by incumbents or Chinese. I do not think this reflects the investment climate in Europe in general. There are fields where startups can do well, and fields where starting something new in Europe is generally a bad idea. I’m not sure that’s always because of regulations.

Cost of layoff is wrong measure of competitiveness, to start with.

Production of innovative technology pushing the frontier of whats possible is the measure. Cost of laying off is one possible non exhaustive diagnosis as to why it might not exist in Europe.

Nobody is going to build any software in US when nearshore is 2-3 times cheaper. American engineers are overpriced and not competitive on the global market.

My claim is never about lack of brainpower in the EU. Clearly theres talent. Its a bit cultural, yes, but mostly its part of a framework that does not reward innovation. For a place with the best car manufacturing in the world and such a drive towards renewables, youd think Tesla should have come from EU. Or even should be replicated there. Why wont it?

If you want dirt cheap contractors who largely accomplish nothing, the US is indeed the wrong place.

You are of too high opinion about Americans thinking that there’s no comparable talent elsewhere.

Didn’t say that

Nobody wants “dirt cheap contractors who accomplish nothing” literally. That’s the figure of speech that has only one meaning in the context of conversation: that you supposedly think that outside of USA, people have lower qualifications. If you intended to say something different, feel free to elaborate.

That’s not how logic works

> Nobody wants “dirt cheap contractors who accomplish nothing” literally

Yes they do. This is why cognizant, Accenture, wipro, etc. exist

Comments here somehow missing the fact that EU alternative to Starlink has already been work in progress for a while with several competitors in the game and billions € of investment. The replacement goes on, the smell of money in the air is felt quite well, digital and technology sovereignty is a popular topic on startup events.

So... the free market is fixing this. lul

I'm on EuroDNS as domain registrar for many years, parked there a couple of companies where I worked. They don't have scammy vibe of other registrars, have multi-user accounts, which is convenient for businesses.

This page is inaccurate and forgets to mention the one of the largest immigration waves - repatriation of Germans from Eastern European countries. A lot of people born in Poland, Czechia, Ukraine, Russia or Kazakhstan are ethnic Germans that were either forced to move (Eastern Prussia) or were invited to come in admission of what they suffered as a result of Nazi invasion to those countries.

Let‘s not underestimate the corruption within Russian intelligence services, that resulted in invasion without any meaningful military planning. They spent the budgets corrupting Ukrainian elites and brainwashing population of the Eastern Ukraine, probably taking a cut in their own pockets, but they were not expecting to be held accountable for the result. The success of Ukraine in early days was in the element of surprise. Their resistance was not expected by Russian leadership, and the invasion itself was not expected by the middle ranks of Russian military. It did not go according to the plan, but plans change, and this war demonstrated weaknesses of all major powers - in a way, Russia did tick some boxes. As successful Ukrainians were in that moment, NATO has lost the war for them failing to mobilize in time, also taken by the surprise (as if CIA weren’t warning months ahead that this is going to happen). If Patriots, tanks and ammo would start flowing in the first weeks, maybe we would see peace in Istanbul accords in 2022 already.

Easy to attack, fast to replace. Every attack requires some planning, a drone and time window. That means attack budget. Spending it on traffic lights does not make sense.