mafribe

Exactly.

Taiwan has spent the approx 120 years on a very different political, economic, cultural track from the mainland. Taiwan diverged from the other subject of the Qing dynasty before Han nationalists began their century long project to forge a united Chinese nation. In particular, Taiwan did not go through decades of communist terror, but did experience the fruit of democracy.

Could you give some evidence for this claim? Here is some counterevidence: [1] says that the top buyers of Russian energy include:

• Hungary: 416 million euros ($488m)

• Slovakia: 275 million euros ($323m)

• France: 157 million euros ($184m)

• Netherlands: 65 million euros ($76m)

• Belgium: 64 million euros ($75m)

[2] suggests that China and India are the main buyers. I don't how reliable those sources are. There is also the problem of how to classify 'laundered' oil that was bought and resold by, e.g. India.

[1] https://www.aljazeera.com/news/2025/10/3/how-much-of-europes...

[2] https://energyandcleanair.org/june-2025-monthly-analysis-of-...

> How the parameters are passed must be decided by the compiler, not by the programmer.

Consider the program

   def f(x):
      return 17

If omega is a non-terminating expression then the call

  f(omega)

might or might not terminate, depending whether the compiler uses a lazy or eager evaluation strategy. This freedom is not something I would recommend to a programming language designer.

The history of exceptions is discussed in [1]. I am not a historian, but think that Milner's ML was the first language with a type-safe exception mechanism. [2] Discusses, among many other things, Lisp's relation with exceptions.

[1] B. G. Ryder, M. L. Soffa, Influences on the Design of Exception Handling. https://dl.acm.org/doi/10.1145/885638.885644

[2] G. L. Steele Jr, R. P. Gabriel, The Evolution of Lisp. https://www.dreamsongs.com/Files/HOPL2-Uncut.pdf

One might decide to act out one's entrepreneurial spirit in a different country ...

TikTok, YMTC, DJI, DeekSeek, ZTE, Huawei, SMIC, BYD use ``slave labour''?

Have you been at recent top conferences in computer science? Every year the number of top papers coming from China increases. A direct consequence of the sustained investment into education. China also leads most European in EV adoption.

You man know Kübler-Ross's Five Stages Of Grief [1]. Like most Europeans, you seem to be stuck somewhere between denial and anger regarding European relative decline vis-a-vis China. I suggest flying over there and taking a look for yourself.

[1] https://en.wikipedia.org/wiki/Five_stages_of_grief

Samsung is not exactly averse to IP theft eg [1, 2] ... If you are responsible for 22.4% of South Korea's GDP [3], one may wonder if we even hear about all other cases ...

[1] https://www.theregister.com/2015/08/25/tsmc_samsung_espionag...

[2] https://www.reuters.com/legal/samsung-hit-with-303-mln-jury-...

[3] https://www.statista.com/statistics/1314374/south-korea-sams...

Country C's regulators can sanction the companies entities in C including denying access to the market in C. (C.f. how the US got Swiss banks to comply with financial transparency laws.)

It would not matter if C == Lichtenstein, but China is currently the world's biggest semi-conductor market ...

The pipeline stages (say: fetch, decode, execute, memory access, register write back), are organised "parallel in space" as transistors on chip. The point of having a pipeline is so the stages can execute "parallel in time".

More generally, parallel in space is interesting because it is a necessary precondition for parallel in time.

Can you explain why starting a war (still ongoing), killing >10k people, and converting Africa's best functioning and richest country into one of the world's worst functioning places is positive outcome? I don't understand this.

The Syrian Civil war was clearly (in parts) engineered by the west. Here is some evidence.

- Western government spokesperson in 2003: https://wikileaks.org/clinton-emails/emailid/18328

- In 2014, the West officially intervened in the Syrian civil war: https://en.wikipedia.org/wiki/US_intervention_in_the_Syrian_...

- Western government spokesperson in 2018: https://www.washingtonpost.com/news/global-opinions/wp/2018/...

- As of 2024 the West still has at least 1000 military personnel in Syria: https://theconversation.com/us-military-presence-in-syria-ca...

Neuromorphic has been an ongoing failure (for general purpose processors or even AI accelerators), ever since Carver Mead introduced (and quickly abandoned them) them nearly half a century ago. Bill Dally (NVidia CTO) concurs: "I keep getting those calls from those people who claim they are doing neuromorphic computing and they claim there is something magical about it because it's the way that the brain works ... but it's truly more like building an airplane by putting feathers on it and flapping with the wings!" From: Hardware for Deep Learning, HotChips 2023 keynote.

We have NO idea how the brain produces intelligence, and as long as that doesn't change, "neuromorphic" is merely a marketing term, like Neurotypical, Neurodivergent, Neurodiverse, Neuroethics, Neuroeconomics, Neuromarketing, Neurolaw, Neurosecurity, Neurotheology, Neuro-Linguistic Programming: the "neuro-" prefix is suggesting a deep scientific insight to fool the audience. There is no hope of us cracking the question of how the human brain produces high-level intelligence in the next decade or so.

Neuromorphic does work for some special purpose applications.

The paper goes out of its way not to compare the 42% figure with anything. Is "42% within the top 5 suggestions" good or bad?

How would an experienced engineer score on the same task?

Note that "some sort of partitioning" isn't a hyperplane. A partition is a set-theoretic concept. A hyperplane is (a generalisation of) a geometric concept, so has much more structure.

This is wrong!

The term hyperplane already assumes that the hypothesis space that your learning algorithm searches has some kind of dimension and is some variant of an Euclidean / vector space (and its generalisations). This is not the case for many forms of ML, for example grammar induction (where the hypothesis space is Chomsky-style grammars) or inductive logic programming (hypothesis space are Prolog (or similar) programs), or, more generally, program synthesis (where programs form the hypothesis space).

Multi-paradigm languages, and there are several popular ones including Scala, need strong software architecture enforcement. The software architects need to decide which style to use, e.g. purely functional and monadic, vs nicer-looking Java-style OO.

What 3.0 debacle?

I've just started a new project and we are using Scala 3. It's wonderful. (Admittedly, it's blank-slate, so no dependencies to previous Scala, e.g. the old Scala approach to meta-programming.)

> expressing how I think.

What you are saying is basically that you like ML-family languages, which also include Haskell, Ocaml, Rust, Kotlin. Increasingly, other languages take the ML-family lessons from 1973 on board!

Hard drug addicts prefer using their limited income for drugs rather than rent. (Note that as of Jun 2023, hard drug addiction has no known cure.)

Without wishing to engage in a discussion of the merits of the respective positions, it might be interesting to know that Maturana rejected Luhmann's use of autopoiesis in social systems.

> Throughout the history of the US

Is the US the world's only country?

Throughout the history of the world, who counts as XYZ has been redefined many times to suit the political goals of the people at the time.

I recall when we stopped being Yugoslavian, and started shooting at each other, as Serbs or Croats. The Russian guy sitting next me married a Ukrainian woman a decade ago. Recently Russians and Ukrainians started shooting at each other. There's a great deal to read on this topic if you're genuinely interested.

“The thing that hath been, it is that which shall be; and that which is done is that which shall be done: and there is no new thing under the sun.”

> money are kept by the people in banks and not released

The banks will reinvest this money. (Cf Fractional-reserve banking).

From the conclusion of the paper: "Extensive testing on formally verified software is necessary for at least two reasons. First, the formal specification may not guarantee all the properties expected by users, but only critical ones (e.g. no miscompilation for CompCert). Second, critical bugs may still remain, because the formal specification might not exactly fit reality. However, in this case, critical bugs are in the—much smaller and simpler—TCB"

> lot of people that use drugs recreationally.

We all know them. HN is full of them.

They are the reason why the production and distribution of drugs is financially viable, why we have functioning markets for illegal drugs. Consequently they, together with the producers and distributors, bear the responsibility for the cost that the existence of this well-lubricated market has on the weak, on those who cannot handle hard drugs recreationally.

Do you think this has something to do with the fact that turnstile jumping has been effectively legalised (in the sense of not being prosecuted) in NY?

Question for you: can you quantify, what fraction of crime and other forms of sociopathy in the NY public transport system you estimate to be committed by passengers who paid their fare? (My estimation: less than 1 percent.)

I don't think it's reasonable to assume that a simple metal gate alone can completely solve complex social dysfunction, a simple metal gate can however help, and, when we refer to turnstile access being desirable, we implicitly assume that we can reasonably expect turnstile use being adhered to, and violations punished with at least moderately high probability.

I did not bring up US public transport as an example of "best practises". I agree that Berlin has a extensive and well-developed public transport, and that is commendable.

> Turnstiles don’t solve homelessness.

Nobody claims they do. My anecdote illustrated the opposite direction: barriers remove one related cluster of reasons, related to personal safety, why some avoid public transport and prefer to drive by car, namely the fear to be accosted by vagrants, pickpockets, and other forms of sociopathy.

Could you justify the term "best practises"?

Last time I took an U-Bahn in Berlin, a guy was urinating in front of me. I have not seen such sociopathic behaviour in public transport in Tokyo, Singapore, Taipei, Shanghai, Hong Kong, Beijing. All are turnstile based. I feel that they are strictly better in almost all dimensions than e.g. Berlin's public transport. In all you pay with some variant of NFC tech, e.g. your phone. Zero effort.

Fine-grained access control also allows for better understanding of train usage, and capacity planning.

Cost of transport is orthogonal to access.

> 1970s

I wonder what fraction of the HN readership gets this joke ...

CNNs are from the 1980s (the "neocognitron" by Kunihiko Fukushima [1]), while the MLP is from 1958 [2]. So nearly half a century, resp. a century old.

[1] K. Fukushima, Neocognitron: A Self-organizing Neural Network Model for a Mechanism of Pattern Recognition Unaffected by Shift in Position.

[2] F. Rosenblatt, The Perceptron: A Probabilistic Model For Information Storage And Organization in the Brain.

Yes, Hinton was on a temporary position at the University of Sussex (IIRC, the Centre for Cognitive Science) for a while, but was not offered a permanent academic position here, when he applied.

SIMD is orthogonal to clock-speed.

There is indeed a lot to gain from having a single instruction trigger more complex behaviour, for example better instruction density, less instruction decoding needed, but all of this is independent of clock frequency.

I think, but am not sure, that the Thinking Machines' CM-2 from 1988 was a 4096 or 8192 wide SIMD machine. Surely, at the time, clockspeeds where low.

Interesting, I had no idea. Wilkes doesn't cite Babbage, oh dear!