midland_trucker

As someone who has dived into Rust over the past few months, the two main reasons for employing it are: A) Bleeding edge performance (Shares the same problem space as C/C++, zero cost abstractions etc.) B) Memory safety (Giving it a rather unique flavour among systems languages - distinguishing it from the above two)

If you primarily work in C#, the latter won't be a problem you will be looking to solve anyway. I would say if the former intrigues you then Rust would be a great place to start.

Nah, my theory is you're seeing run-of-the-mill brash business dealings with the unusual characteristic of being broadcasted publicly.

You wouldn't accuse the same of CEOs from more 'boring' companies because we only get exposed to the results through sterile PR communication, despite the fact the same candid conversations still happen.

Arianespace of course a huge success for a brief period, until it wasn't.

Liquid-fueled setup is largely unsuitable for ICBMs. Want something that can sit in the silo for years and then be immediately available - so solid rocket motors are the way to go, plenty of other defence companies that are specialists in this.

They only just announced it a few weeks ago, surely not launching it would be the business decision if this was the case.

I'm yet to see a study that convinces me of the potential of this concept. All of them when you get down to the details have at least one of:

  - An unrepresentative test set , i.e. a population with similar blood pressure at recording in which merely predicting a constant 'typical' reading would perform well in terms of accuracy.
  - A major methodology flaw, such as 'demonstrating' an ability to detect fluctuations from a baseline calibration cuff reading in situations where you would expect no significant fluctuations to occur.

I don't have access to this particular paper, but suspect it has the same issue owing to the relatively small sample size.

This is my gut feeling about a lot of activity going on in modern corporations. Private equity firms, for example, have been historically blamed for a lot of once great engineering companies in the UK being combined into a bloated mess of activities - which ends up ultimately sacrificing competitiveness.

However I feel like your average voter shaking my fist angrily a bankers sometimes when trying to put it into words (I don't honestly have a lot of insight on how it all works).

Have you got any good book/essay recommendations on over-financialisation?

I recently needed to render markdown to view some notes. Thought opening chrome w/ a plugin or VSCode felt similarly sluggish options to do such a simple task. Now a few weeks after playing with textualize this pops up! Neat tool.

> In Go, I found that using an interface was not free: it can make the code slower. In C++, if you implement the following type and call count on it, you find that optimizing compilers are able to just figure out that they need to return the size of the inner vector.

I'm surprised at this, do Go interfaces really introduce much overhead? Of course this depends on the level of performance you care about but surely, being a statically typed language, lots of the same optimisations are available.

Exactly, it's just the wrong people at the top and the wrong incentives. Is Arianespace structurally capable of producing economical access to space? Probably not.

Perhaps we're talking past each other, is "loss threshold" a specific term in LLM literature?

Merely pointing out that the debate as to whether we are compute or data limited (OP) has not concluded at all; There are lots of compelling theories on relationship between the two.

That's a fair point. Not at all like training on Wikipedia in which nearly every sentence has novelty to it.

Then again it would give you data on every accent in the country, so the holy grail for modelling human speech.

> Dataset size is not relevant to predicting the loss threshold of LLMs. You can keep pushing loss down by using the same sized dataset, but increasingly larger models.

Deepmind and others would disagree with you! No-one really knows in actual fact.

[1] https://www.deepmind.com/publications/an-empirical-analysis-...

The WebGL equivalent sites were a delight to go through; So well structured and thoughtfully displayed. It's so great to find a resource you can put some trust in amongst all the outdated and fragmented code out there.

Looking forward to ploughing through this when I have the free time to.

I certainly wasn't and appreciate you posting it, great site! A little intimidating initially, but after a minute curiosity has you flicking between the different controller channels and stages in the mission.

I'm not sure I can get on board with the sentiment of your first point. Anaesthesia as another example took more than a century to get right, had some huge failures and wasn't taken seriously. It surely earns the label of innovative but just happened on a much larger timescale.

As a programmer not part of this world: what's the go-to alternative in this situation?

Does ECS not naturally arise from using OOP? Trying to think if you are creating instances of players/physics objects/items how that wouldn't turn into a kind of entity system.

Perhaps a lot of people wouldn't agree and I admit I'm stirring up debate a bit to see what other people think.

To me it seems if we see orbital success in the next 5 yrs, mass production of starship thereafter and SpaceX sticking to their Earth-to-Earth transport idea - this seems plausible. Airliners developed about as quick, but then again its a far harder environment - it's nice to dream.

[1] https://www.youtube.com/watch?v=zqE-ultsWt0

(edit rambling on about the wrong system)

A thing I find mind-blowing is that if this concept succeeds, that I and everyone reading this will be able to afford to go to low earth orbit on a relatively modest salary within the next ~20 years (conservatively).

This feels like quite a difficult reality to internalise given we all grew up seeing only a select few exceptional people, with the stars aligning in their career, be able to do that.

Goes to show that nailing down the 'HN crowd' opinion is more difficult than it seems!

I find it really hard to see how productive a collective pause and 'think' about something so inherently unpredictable will be.

> "implement a set of shared safety protocols for advanced AI design and development that are rigorously audited and overseen by independent outside experts"

Who deserves to be called an expert on this? Feels like Economics or something, where you have camps of thought advocating for themselves but little way of knowing who's right. Best to break things and develop antibodies whilst the stakes are still low.

It was a specific rebuttal to the parent comment which seemed to not understand the reality of resources in vs resources out of animal agriculture. Of course you can caveat a lot about my comment, beef for instance is arguably the most inefficient example of meat you could cite (amongst slightly less inefficient alternatives) but illustrates the problem well. That is the nature of compressing such a complex issue into a small HN comment.

> Why is water consumption an issue perse? What did the 5kg of food consist of? Is that 5kg we humans could've ate? I doubt it. Throughout time we've given our animals scraps which we think is beneath us.

Right - the discussion is about land use. We don't typically eat animal feed but allocate an enormous sq footage that could have been used to grow something we would might like to consume directly.

> Take for example a cow in England eating grass and drinking water which directly came from the rain. What is inherently bad about that?

Agree, we have plentiful access to drinking water in the UK and so have enough to allocate to irrigation and all the other water intensive activities involved to do this. Unfortunately on this particular point, the majority of the beef you eat does not come from here. This issue becomes much more pertinent if you live in say California or Spain.

I'm pointing out that you don't seem to realise how insanely resource inefficient meat is:

[1] https://ourworldindata.org/grapher/feed-required-to-produce-... [2] https://cbey.yale.edu/our-stories/disrupting-meat

Napkin maths from the first source suggests 5kg of feed to produce a 200g steak. All while losing all but 4% of the original calories. This is of course ignoring the vast water consumption and emissions (and of course ethical problems) which are issues worth talking about in themselves.

Conclusion is quite the opposite of your original comment, it seems we could feed the world MANY times over using less land on a 'plant based' diet.

> In this case a Plant based Diet would require much more agricultural land for the same calorie intake than a balanced meat and plant one.

Where do you think the calories in meat ultimately comes from?