Which makes me curious about what Marc actually meant. The quote itself raises eyebrows.
EDIT:
From checking in with Claude about his talk.
> So the thing he was arguing against was specifically what he sees as a modern therapeutic culture — the expectation that people should examine their motives, feel guilty about their actions, and look backward. He wasn't framing it as a philosophical position so much as a practical one about founder effectiveness.
Yes, the article is deliberately misconstruing that.
The author says outright in her comments that the aim is not to engage with what Marc actually meant, but to chastise him for bad thoughts expressed on Twitter:
“I get where you’re coming from - and I’m not sure I’d disagree with Marc, if that were the point he made. But his hours // days long crash out on Twitter I think made it clear that a lack of introspection was certainly at work…”
It's not at all a similar analogy, and you know it. China is a threat. I assume you personally don't care about it, but it is in our interests that it does not dominate the AI industry.
China is a threat, if you live in Taiwan. They lack the ability to project force any further than that. You pin down IR guys on this issue they eventually relent and suggest actually the biggest issue with China is that their claimed zone of control includes a lot of ocean trade routes. But honestly, Xi has been better for international trade than any US admin in the last 15 years, I don't see them doing anything but protecting it. They are like 100 times less threatening than the USA, and that was before president tantrum, who has acted (quite recently) to destabilize global trade.
That doesn't work when your opponents pray for death and see martyrdom as victory.
This is genuinely how Shia extremists think. They have nothing to lose and will sacrifice everything and everyone for their cause. They don't care about Iran or Iranians or prosperity of the nation.
Every country that has a opposition diaspora says the same stuff you're saying here. For what is worth, you could be from a family of Savak secret police members.
It is kind of funny, and I am not a muslim, but I am curious enough about history of religion to get absolutely baffled by this demonization of Shia.
Shia is actually way more moderate and compatible with western values. Most terror attacks in the west actually are linked to wahabbism (a more radical sunni variant) than to Shia Islam.
No, he or she is saying that even Americans who have moved overseas could be heard to complain about the "fascist" authoritarians in power in the US now. They would sound functionally identical to an Iranian emigrant talking about Iran; only the details would differ
More to the point. Presume Trump cancelled elections and became a dictator. Then a popular revolt overthrows the MAGA dictatorship, starts persecuting MAGA bureaucrats and leaders. Like in any revolution excesses would happen, the economy would, at least temporarily, take a nose-dive, basic services would stop, and so on.
In such a situation, lots of people would presumably leave the US to form a diaspora. Some of those of course, would have been MAGA people directly culpable in the former illegal power grab by trump.
The Sha was not a loved wise leader, he was also a brutal dictator who directed a Comprador elite at the expense of the majority of the persian people. Some of the Iranian exilees just want to go back to act as colonial administrator for the western world like they did before the revolution.
Even if you consider the Islamic Republic evil, you need to be careful before enthusiastically buying a narrative from one side, because a lot of times politics is just the eternal fight of evil against evil.
That's a psychotic thing to say about starting new wars and aiding genocides. The only thing that's being defended are the profits of western oligarchs.
A good lawyer could argue that Tesla must be aware that exiting the auto market would immediately crash the value of all existing Teslas because it would essentially create a sunset date for those vehicles, given how much software they're running. Good luck to anyone trying to sell a used Tesla once that announcement is made, because who would buy a car that is going to be bricked at some point?
You should check out Michael Levin. Cancerous cells do not grow organ-like structures. Normal cells communicate with other cells as a network to control growth.
Thanks! Currently focused on 2D charts. That's where the "big data" performance problem is most painful.
3D is coming (it's the same rendering pipeline), but I'd want to get the 2D story solid first before expanding scope.
The slice animation is doable though - we already have animation infrastructure for transitions. An "explode slice on click" effect would be a fun addition to the pie/donut charts.
What's your use case? Dashboard visuals or something else?
Lattner is a smart guy, but I think Mojo might be the wrong direction.
Time will tell.
History has not so far been kind to projects which attempt to supplant cPython, whether they are other Python variants such as PyPy, or other languages such as julia.
Python has a lot of detractors, but (despite some huge missteps with the 2-3 transition) the core team keeps churning out stuff that people want to use.
Mojo is being positioned "as a member of the Python family" but, like Pyrex/Cython, it has special syntax, and even worse, the calling convention is both different than Python, and depends on the type of variable being passed. And the introspection is completely missing.
Honestly, I feel like Julia might as well beat Mojo or sommat to the punch, sooner or later. It has some facilities and supporting infrastructure for a lot of scientific and data-handling tasks surrounding ML, if not for compiling and dispatching kernels (where XLA reins supreme to anything in the CUDA ecosystem!) For example, Bayesian programming like Turing.jl is virtually unmatched in Python. It's been a while since I looked at Lux.jl for XLA integration, but I reckon it could be incredibly useful. As long as LLM's and RLVR training thereof should continue to improve, we may be able to translate loads of exiting Pytorch code eventually.
But the first thing that gave me pause about Julia? They sort of pivoted to say "we're general purpose" but the whole index-starting-at-one thing really belies that -- these days, that's pretty much the province of specialty languages.
You're not supposed to admit it, but I never cared for Dijkstra's arguments on the matter. The same goes for his GOTO tirade, although that has been distorted by time somewhat. Pascal is using 1-ord, Fortran, R, Mathematica. If anything, it seems there's a longer tradition of 1-ord in scientific computing. In this view, I must agree insofar I don't think Julia people are serious about their "general purpose" stance whatsoever. But hey, these are merely idiosyncrasies. People say multiple dispatch is the shit, but it's just one bit of the puzzle with Julia. How they managed to engineer a solid foundation, semantics like that, without unnecessarily sacrificing performance—I don't think they get enough credit for that from programming guys.
> I never cared for Dijkstra's arguments on the matter.
It resonates with me, as does little-endian. There are evergreen arguments about both of these, but they stand out to me as two of the few places where I appear to be in sync with the majority.
> Pascal is using 1-ord, Fortran, R, Mathematica. If anything, it seems there's a longer tradition of 1-ord in scientific computing.
Actually, Pascal let you define the starting point, but, historically, COBOL (business) and FORTRAN started at one, and these days, it's Matlab (engineering), as well as R, Mathematica, and Julia. But, really, my point was that Julia is hyped to take over from Python, and it seems if that's the focus, then you should think long and hard about what features of Python to change, and, well, obviously that wasn't really the direction they came from.
> semantics like that, without unnecessarily sacrificing performance—I don't think they get enough credit for that from programming guys.
It's good work, but at the same time, a bit overblown IMO. For example, the creators claim homoiconicity, but that stretches the definition of that word to where Python would be homoiconic as well, and it's obviously not.
As far as the multiple dispatch goes, that's actually not that difficult of a puzzle piece when you are compiling ahead of time with unboxed objects.
The fact that it displays as 1 + 2 * 3 instead of (+ 1 (* 2 3)) ... meh. The whole point is to make metaprogramming easy, which Julia is quite successful at (for better or worse -- some people shouldn't be allowed near macros).
This 100% concerning Julia. Until recently at least, there was a ton of fundamental correctness issues. The issues may have been fixed, but it certainly does not inspire trust.
Honestly, I'm not familiar with it. I had only played with RxInfer, if only to try the so-called "message-passing" paradigm. My grasp on probability is really lacking, in fact I picked up prediction markets and Julia to get better at it.
If you don't mind me asking, what's the deal with NumPyro why you chose to work it?
Well I did my PhD working with Probabilistic Torch, then shifted to Pyro for most applied probabilistic programming in PyTorch, and then wanted to learn Jax. So I moved over to Numpyro.
I don't think they intend to supplant cPython, but to make it easier for someone to take a piece of software written in high-productivity Python and make it more performant on specialized hardware by using a similar syntax.
Long term it can find applications outside of ML-specific hardware.
You know, 1.5x for overtime seems great until it is enforced for everybody (like it's done in Quebec) and so it is forbidden by company leadership unless it is deemed a necessity. So for people with more time on their hands than responsabilities, they usually can't (voluntarily) work more in their main job to boost their pay, and so they get side jobs, which are usually not declared.
Mandated 1.5x overtime has consequences.
Another side effect (that I know happens in manufacturing jobs) is that people will deliberately slow their work during the week to work overtime on the weekend. I'd wager this is very common in jobs with frequent overtime opportunities.
> So for people with more time on their hands than responsabilities
What kind of widget are we building where underemployed meth-heads get to show up for as long as they want and set their own pay schedules? Nothing in life works like this.
Look at it this way: Why would you hire someone who doesn't think they're worth market rate?
EDIT: From checking in with Claude about his talk.
> So the thing he was arguing against was specifically what he sees as a modern therapeutic culture — the expectation that people should examine their motives, feel guilty about their actions, and look backward. He wasn't framing it as a philosophical position so much as a practical one about founder effectiveness.
https://claude.ai/share/9c5611f7-fd0e-4f76-bd39-e1129c035a4f
reply