HN Theater @HNTheaterMonth

Maybe I'm a particularly dull engineer, but I've taken several aspects of personal advice from what he has said in interviews (the especially technical ones, not the ones aimed at a mass audience where he repeats his standard canned speech) and found them useful for myself personally.

Here's two examples I've found particularly insightful that shows he has some ability to talk about engineering details.

This example where he talks about the choice of steel for Starship as opposed to any other metal, something that would be an otherwise unsual choice: https://youtu.be/vLC5W53Fsyg?t=936

This example that I've personally incorporated into my own thinking where he talks about his "five step process" for engineering design refinement (watch at least until he starts talking about Tesla Model 3 battery stuff): https://youtu.be/t705r8ICkRw?t=805

⬐ NaturalPhallacy

>This example that I've personally incorporated into my own thinking where he talks about his "five step process" for engineering design refinement (watch at least until he starts talking about Tesla Model 3 battery stuff): https://youtu.be/t705r8ICkRw?t=805

I knew what you were talking about when I started reading your comment. "Make your requirements less dumb" first seems so obvious once you've learned it.

All the denigration directed at him seems to come from people who've only read headlines about him from sources who hate him.

Listening to Elon Musk talk about the ideal manufacturing process as "A Genie turning your Raw Material into Rocket Engines" ...and then comparing that vs how much time and money Space X wastes hammer sheets of metal into shape https://m.youtube.com/watch?v=t705r8ICkRw

I saw a tweet that a twitter home-page visit is 100 poorly-batched RPC calls, and I believe it Not many bytes are necessary to deliver some text and images and a css layout

⬐ onion2k

Not many bytes are necessary to deliver some text and images and a css layout

To be fair though, if that's what you think Twitter is then you need to think about it some more.

I mean, I don't think the dumb tunnel under vegas is good evidence of not being a brilliant engineer. He obviously has a very deep understanding of engineering type shit as well as engineering management[0][1]. He's got engineer brain! Engineer brain can make you do a lot of really stupid shit even if you're a great engineer.

That being said I absolutely don't think he's playing 10D chess, he's got a few big Ws and it's gone to his head in a disastrous way. He can be a brilliant engineer/engineering manager and a total fucking idiot at the same time.

[0]https://www.youtube.com/watch?v=YAtLTLiqNwg [1]https://www.youtube.com/watch?v=t705r8ICkRw

⬐ anonyme-honteux

To be honest the thing that worries me most about the dumb Las Vegas tunnel, is not that he had a bad idea. I have bad ideas all the time too. But my bad ideas don't turn into dumb tunnels because I have limited resources (he doesn't) and because I have feedback from the harsh reality.

The dumb tunnel makes me think Elon Musk is fully insulated from reality. Nobody around him dared to tell him the tunnel was dumb. And or he didn't listened.

Fast forward today where he decides on a whim that every engineer must stop working remotely and must instead work like crazy to satisfy his ego. And here again the feedback from reality seems minimal on him.

I think and I hope that the good engineers at Twitter are making plans to leave this terrible boss ASAP.

> his success is mostly throwing money at eager (and smart) engineers

This meme is on fire but I did not expect it here on HN where people can understand what he is talking about. There are many videos like this [1] and interviews (e.g. with Munro regarding Tesla) where the guy shows deep technical knowledge.

Besides, try making a startup throwing money and smart engineers. Just money will not even get you the really good ones. People who have a choice don't work for morons.

I mean, sure, there's plenty to hate (there was much more to hate about say Columbus, Gandhi or Ford), but if we were dismissing achievements of people who give others good reasons to hate them we wouldn't have calculus.

Let's try to let people stand out without gaslighting them if they are doing some great things. There are many people who smell real bad, are a*holes, have really weird opinions on some topics and were necessary for us to be able to write these comments here.

1. https://www.youtube.com/watch?v=t705r8ICkRw

⬐ Salgat

He doesn't have deep knowledge, he a broad high level understanding of the technologies and processes used at his company. In the interview you can see how he's aware of what the major components are supposed to do, but he'd be completely useless if you told him to do the engineering design and analysis behind it. Which is perfectly fine, that's how it's supposed to be as a CEO.

⬐ intothemild

Did you just compare Musk to Ghandi and Ford?

⬐ comboy

Don't forget Newton. Not sure what is the problem, Musk achievements not big enough to be a dick?

You can talk about multiple dimensions of these people, I'm trying to show something between "did something worthy" and "was really not a great guy" axes, and you seem to have a problem that they are not even close on "historical figure" axis, I'm guessing. Their hair could also be way different.

I really do think it's the media you're consuming.

Here he thanks Tom Mueller: https://twitter.com/elonmusk/status/1333544636137283586

Here he thanks some Tesla club members: https://twitter.com/elonmusk/status/1211836235762880512

Here he thanks NASA for all the support over the years, "without which this would not have happened": https://twitter.com/elonmusk/status/1219127317957615617

Here he says that "Gwynne is the best": https://twitter.com/elonmusk/status/1538400836669296640

That's just a random collection I found with a couple of quick google searches.

If you have actually watched for example the starbase tours with Elon, it's difficult to claim that he isn't "involved": https://www.youtube.com/watch?v=t705r8ICkRw

⬐ bigmattystyles

I didn't mean to say he wasn't involved at all, just that imho (which doesn't matter) most of these companies would be fine without him. He is a savvy business player, knows how to grab available resources as well as the spotlight. I don't see anyone heralding Bezos or a few other CEOs I could name to this level and he's quite a bit more successful.

Sidenote, all but a few of his twitter mentions of Shotwell are responses to mentions of her and it seems like he's injecting himself into the discussion.

⬐ dwaltrip

Bezos famously has been trying to build orbital rockets for two decades and hasn't yet succeeded.

So there must be some difference between them on this point.

If I had to choose between network television and YouTube, I would choose YouTube every day of the week and twice on Sunday. I mean, I enjoy Seinfeld too, but on the whole YouTube is both more entertaining and vastly more useful and educational than network television. Not because the content has the highest production values, but because the content matches my interests and needs. That's the true power of targeted advertising. It allows creators to focus on niches that can't be served by content produced for a mass audience. Even large 'niches' get dumbed down to the lowest common denominator for network television.

Case in point: Here's Jay Leno interviewing Elon Musk about SpaceX Starship for network television [1] vs YouTuber Tim Dodd [2]. Leno's program is fine for people who don't follow space. For my niche Dodd is vastly better, but he can't draw a wide enough audience to make money on non-targeted ads like Leno can.

And the thing is, everyone has their niches. There is no large population of people out there who are average in every way and need the network television version of everything. The average person doesn't exist.

[1] Leno: https://www.youtube.com/watch?v=wluBlr1j4qk

[2] Dodd: https://www.youtube.com/watch?v=t705r8ICkRw

⬐ Retric

Dodd got 6 million people watching a 1 hour segment. That valuable even without advertising targeting beyond estimates of viewership based on content.

Now, you can quibble about how much targeting increase revenue but the funny thing about total advertising is it’s fairly consistent through time. Tracking failed to significantly increase worldwide advertising spend, it only redirects it.

⬐ modeless

6 million views (a view only requires 30 seconds of watch time, not the full hour) over a period of 14 months, for the most viewed video ever on his channel. If he could draw that for every video, maybe he could come close to competing with Leno. But he can't. I mean, looking at YouTube numbers alone, Leno's show's most viewed video has 34 million views. And of course the vast majority of his viewership and revenue is elsewhere.

Dodd simply couldn't have a network television show with this kind of content. It wouldn't come close to making sense. If he did a network television show it would have to be a lot different, almost certainly in a direction that would make it less interesting for me personally.

My argument has nothing to do with global ad spend, that's a red herring. The key is how those ad dollars get distributed. Targeted advertising makes niche content viable in a way that non-targeted advertising can't match.

⬐ Retric

You can have non targeted advertising on niche content. In fact YouTube will display advertising to any user even those it knows nothing about.

The only difference is relative pricing, but again the global population and global advertising spend doesn’t depend on tracking so if it was banned little would actually change.

⬐ modeless

> global advertising spend doesn’t depend on tracking so if it was banned little would actually change

> The only difference is relative pricing

... the relative pricing difference is the whole difference I'm talking about. Without targeted advertising niche content makes less money, general audience content makes more. So you agree with me that this is true. Then you say that this is a small change. I say it has enormous effects on the content that gets produced. Honestly, this is transparently obvious.

⬐ Retric

It’s not that simple. Tracking doesn’t universally increase spend to all niche content, it reduces spend to some niche content and increases it to other niche content. So there would be some changes to which niche content is created but it’s not going away.

Just look at how much YouTube content is sponsored via an Ad inside the video. That isn’t tied to your personal history.

⬐ modeless

Certainly I wouldn't say that every single niche benefits from targeted advertising without exception. There are exceptions to every rule. But the vast, overwhelming majority of niche content would be worse off without targeted advertising, and this would overall have a strongly negative effect on the diversity of content produced.

⬐ Retric

Ok that I disagree with, if anything niche content in a world without tracking should attract a higher percentage of advertising.

If I manufacture say jigsaw puzzles and I can’t target individuals then I want to link to show up on a YouTuber with puzzle related content. But with tracking I shouldn’t care about the videos content as much and should be happy to show up on cat videos as long as tracking supports the association. That same logic would seemingly extend to any activity with associated products which is basically everything.

On the other hand soda manufactures are presumably less picky and could advertise on both.

⬐ modeless

Advertising soda is always worth it without targeting. But advertising jigsaw puzzles might not be. That jigsaw marketing money goes elsewhere, not YouTube, and YouTube jigsaw people don't have a lot of ads for their videos, because nobody else is choosing to specifically advertise on jigsaw videos. Maybe even if jigsaw people do advertise, they don't have a ton of money to spend on ads, so jigsaw content is still hurting. But with targeting, there are effective ads on every video, and content matters more than ad spend targeted at your niche. People choose the content they like by watching it, instead of advertisers choosing the content they like by steering their spending. Because there are far more people than there are advertisers, this promotes more diverse content.

We could continue to make up just-so stories about the effects of targeting on niche content. But I doubt we would convince each other. Is there any research on this topic?

⬐ Retric

Do you actually find that reasoning convincing or are you just trying to make a point?

Watch this walk around Starbase with Elon and Tim Dodd, and you'll get a pretty good sense that he has an engineering brain that really understands the complexities of design, build and execute more than anything. https://youtu.be/t705r8ICkRw

⬐ bmitc

I will watch it, but is he the same the “engineering” brain that thinks his Tesla tunnels are a good idea, continually promises tech that is decades away from realization is right around the corner (like self-driving cars or a humanoid robot or a Mars mission), and got replaced as the runner of the actual engineering organization (SpaceX)?

⬐ towaway15463

I’ll take a loose lipped dreamer that eventually delivers on most of his promises in a spectacular way over a buttoned down PR rep that consistently goes over budget, pushes delivery dates by years and drops a disappointing and outdated product at the end any day.

⬐ martyvis

Elon clearly overpromises - but he also over delivers. SpaceX's current capability to deliver payloads to orbit is so far ahead of anyone else (particular considering they started from scratch only 20 years ago). Even their Starlink in just a few years had leapfrogged over incumbent providers. Tesla may not have achieved all they have promised but there is no denying they are the company that is leading innovation in not only EV but even in their approach to manufacturing.

⬐ camdat

> but he also over delivers

My family has had reservations for the Roadster, the Cybertruck, and we have a first-edition Model 3 & X.

Tesla has overpromised on all of our reservations, and the final product, imo, is not "overdelivering". It's honestly just barely "delivering".

>but even in their approach to manufacturing.

This is pure Tesla propaganda. Yes they have built bigger versions of existing technology (supepress, gigafactory), but they're definitely not leading innovation in car production, as evidenced by M3s still having the largest panel gaps in the industry.

⬐ kcplate

Well, I am not an Elon Musk fan or hater and I don’t own any Tesla product. However, I think if your standard for him is “he hasn’t delivered yet on the two brand new proposed models of high-priced EVs that my family has reservations to buy”—I can tell you that I can’t drive 2 minutes without seeing several Teslas, I can see strings of Starlink satellites sometimes when I look into the sky, and at least a couple of times a month I can see SpaceX launches from my backyard.

Seems to me Musk may not be delivering on what you want, which I suggest is a couple of items serving a tiny elite market, but no doubt he is delivering something.

Here you have Dr. Daniel Rasky, a NASA Senior scientist talking about SpaceX and Elon. Daniel Rasky was involved in specific studies that were comparing operation of NASA and private companies. He also has first hand experience from working at SpaceX and with Elon. He is expert on advanced entry systems and thermal protection materials and he was for example involved in decision to select PICA as a heat shield for Dragon, and he explains how that went with Elon.

https://www.youtube.com/watch?v=MxIiiwD9C0E

Here is Elon answering questions of National Academy of Science members about Starship. He explains in detail why he pushed the team towards steel and away from carbon fibers:

https://youtu.be/da3iF2Np51A?t=951

Here you have another direct conversation with Elon:

https://youtu.be/t705r8ICkRw?t=172

Here is Elon discussing with Sundy Munro mass manufacturing and engineering.

Sandy Munro is a veteran engineer from General Motors supplier and directly at Ford. Sandy Munro is now running a consultancy company that tear downs cars and suggests improvements to wide range of clients from the car industry.

https://youtu.be/YAtLTLiqNwg?t=78

⬐ coffeeblack

Thanks, half of the videos I didn’t even know.

Also, Musk was doing a physics PhD in battery tech at Stanford when he dropped out to found Zip2.

⬐ Isinlor

Yup, I think he has done like 1 day of it, before he dropped off.

There is some long running pattern here with getting cold feet on bad decisions ^^ .

⬐ coffeeblack

He was accepted into it, that’s the important part. But then the internet took off and it was the right time to be in that place. Remember, back then he was just another immigrant kid.

It’s funny how quickly the left changed from loving him for turning the car industry on its head to hating him for taking away the left’s hate machine. Oh well.

⬐ ceejayoz

> He was accepted into it, that’s the important part.

Being accepted is not the important part of PhD work, no.

⬐ coffeeblack

Am I on Reddit here?

⬐ meatsauce

Don't be jealous of Musks intellect. The behavior is very childish. Just study and try harder and one day you can be successful, as well.

You need to watch this interview with Tim Dodd (the space YouTuber) and you'll realise in the first 15 minutes he really is an engineer https://youtu.be/t705r8ICkRw

If you watch any of the starbase videos (part 1: https://www.youtube.com/watch?v=t705r8ICkRw)

It's clear he knows a thing or two about rockets. Obv no single person is designing the thing, but it's clear he's highly involved.

⬐ kevinsundar

I was going to link these videos too. Its evident he's closely involved with technical discussions and decisions.

A CNN 2 minute video chopping up what he says to make it sound more grand is not him being a "showman". It's called video editing. You can tell the video has been HIGHLY edited because it's not how he talks.

> it's really hard to believe when he's put out stuff like this.

https://www.youtube.com/watch?v=t705r8ICkRw?t=122

https://www.youtube.com/watch?v=SA8ZBJWo73E

https://www.youtube.com/watch?v=9Zlnbs-NBUI

This is how Musk actually talks.

> Like, did he use up all of his thinking fuel on designing rockets that he didn't understand why the boring company's mission was a total non-starter?

He doesn't spend a lot of time working on the boring company to be frank. It's a small side project. He's doing it to see if it's possible.

⬐ skulk

The point is, building tunnels underground for personal transport vehicles as a "solution" to traffic is a terrible idea, and something only a showman can sell. Even if he has an awkward 2 hour speech about it, he succeeded as a showman because Vegas funded this terrible idea (just look how it ended up).

⬐ mlindner

> The point is, building tunnels underground for personal transport vehicles as a "solution" to traffic is a terrible idea, and something only a showman can sell.

The fundamental point of boring company is to reduce the cost of underground boring. Shoving cars through tunnels is a means to an end of reducing the cost of underground boring. It's about finding a path to make some kind, any kind, of money off of boring in order to make the product development costs self-feeding. This is why the idea of cars driving through the tunnels was done rather than the original idea of putting cars on platforms through the tunnels. Cheaper and faster to market.

Musk started boring company with the idea that transportation needs to be 3 dimensional, and not just adding another single layer of transport under the ground (trains/subways), but having a whole ton of layers going deeper and deeper into the ground. Now they've done their first iteration of tunnel, which is quite bad, rinse and repeat until it gets cheaper.

BTW, Musk has never come out against subways, his point was just that in the US, people drive into cities in cars, so if they're using cars, then the transport system in the city needs to handle cars in a three dimensional way.

⬐ hef19898

So, like subways? Or every single road and train tunnel anywhere on the planet so far?

Moving transportation underground is such revolutionary idea that it exists for a century by now.

⬐ raspberry1337

>Moving transportation underground is such revolutionary idea that it exists for a century by now.

You ignored the essential part of his well written and detailed explanation - its not about the fact that its underground, but the cost of creating underground pathways, for cars or whatever. It saddens me that bad posts like yours infect HN.

⬐ hef19898

It saddens me tgat at HN people start to blindly believe certain people's claims without checking precendce. In Musks case based on, IMHO, pure hype and his celebrity status. So I guess, same?

⬐ raspberry1337

> In Musks case based on, IMHO, pure hype and his celebrity status

Or his proven track record of creating groundbreaking companies such as SpaceX and Tesla, which has made him the worlds richest man and why we are talking about him right now? Jeez

⬐ fakethenews2022

There are differences. I don't know if the idea will actually work but here is what I understand of it. Use an analogy of cities as computer chips/networks.

1. He wants to move roads to 3d just like computer chips are 3d. This should increase bandwidth.

2. He wants to packet switch the system instead of batch. This should lower latency.

3. For the first two to work, you need to greatly reduce the cost/speed to construct them. Think Moore's law.

4. One way to do this is to speed up the boring machines.

5. Another way is to lower the additional necessary equipment. Tunnels carrying internal combustion cars need air circulation lest people die from carbon monoxide poisoning/ lack of oxygen. EVs don't need that.

6. Smaller tunnels are much faster to construct so constructing one for a subway would be much slower than one for EVs.

Arguments against this can either attack the premise that cities are necessary (remote work), that bigger cities are necessary (why not have smaller cities), that existing or other transportation can handle it better, or one or more of the above chains of thought are flawed.

This line of thinking is also found in his latest factories where the line is just a line of cars moved by carriers that aren't physically anchored to the floor). He makes an analogy of factories as computer chips.

⬐ dwighttk

>Tunnels carrying internal combustion cars need air circulation lest people die from carbon monoxide poisoning/ lack of oxygen. EVs don't need that.

Maybe not as much, but you’re still gonna need air circulation

⬐ hef19898

You lost me with comparing transportation infrastructure to micro processors and apllyin Moore's law to infrastructure work.

⬐ benibela

And chips are not even 3d, are they?

At most there are memory chips with just a few layers

⬐ bombcar

Even if nothing else, a boring tunnel could be used for “remote parking garages” and if the tunnel robot takes your car ten miles away to store it what would you care as long as it was available when needed?

⬐ ygjb

You are really committed to your Musk apologist bit, finding any possible edge case that might justify a clearly bad idea...

⬐ hef19898

So Musk moved the Monorail underground... Isn't Mars' atmosphere already a hyperloop?

⬐ xorcist

It's not intended as a solution to traffic congestion. It is intended to give local politicians new talking points to delay any large scale investments in public transportation.

Who knows how we'll solve transport in the future? Maybe we'll all be in personal cars underground and any massive investments in rail will be for nothing?

It's a variant of the general idea of pushing things like elevated monorails that you know are so expensive they won't be built in the near future. Or, indeed, very low pressure tube transportation systems.

Car companies have done this type of lobbying work since forever. Just as the oil companies when they push for more production, but since all oil will be biofuel in the future, in the end it will be a net positive for the environment.

Falcon 9's safety record would indicate that it is possible to have development cycles that push boundaries and maintain a high reliability production system. With regard to Musk's statements about the relationship between innovation and failure, below is a quote from the linked video about optimization in the development process, a quote that provides a more nuanced view than the pithy quote.

Then, try very hard to delete a part or process. This is actually very important. If you’re not occasionally adding things back in, you’re not deleting enough. The bias tends to be very strongly towards “Let’s add this part or process step in case we need it.” But you can basically make “in case” arguments for so many things. And for a rocket that’s trying to achieve, trying to be the first fully reusable rocket…you have to run at tight margins because if you don’t run tight margins you get nothing to orbit.

https://www.youtube.com/watch?v=t705r8ICkRw

His target audience is not the highly technical, but the public at large - including children. He's a space ambassador in that he wants to get the current and upcoming generation excited and learning about space.

With that said, he has been able to get in-situ interviews with rocket company CEOs including his amazing 2 hour almost single take video tour with Elon Musk [0] where I was able see Elon comfortable in speaking much more deeply, candidly, and enjoyably than with any interviewer/reporter prior. Tim Dodd is no professional engineer but he really knows his stuff. His personality helps bridge the gap between the highly technical and the layman.

[0] https://youtu.be/t705r8ICkRw

⬐ mc32

That’s a good point. I’m not an engineer yet I prefer there be more description and less personality shining through; but you make a good point about target audience. And as you say he prepares for the material and does well to understand the experts.

⬐ an9n

Well sure, but there are moments where Elon just flat out ignores him, so idk.

⬐ tnorthcutt

Well put.

Personally I love that Tim's personality shines through in his videos. His enthusiasm is infectious and just plain nice to see, IMO.

⬐ NikolaeVarius

Why. The personality removes so much information density.

⬐ panick21_

Then just read a university text-book and don't look for that in a educational video.

⬐ baq

Content can be also be entertaining to watch. Not everyone wants to superoptimize their time.

⬐ kataklasm

Well, after spending eight hours working demanding jobs, be it physical or mentally, everyday people usually like to spend their free time with something a little more relaxing, and for most people that isn't reading documentation or reading highky technical engineering papers but rather watching a highly enthusiastic guy try and share his joy of rocket science.

Progress at Starbase, SpaceX's base in Boca Chica, TX is done at an astonishing pace and very impressive.

Some of the stuff seen in the video series from August by Everyday Astronaut has changed already.

https://www.youtube.com/watch?v=t705r8ICkRw

> This is objectively false.

I commend the response by user panick21 and would add that you can watch Elon Musk demonstrate his acumen in this three part interview on Everyday Astronaut. You can't fake this kind of breadth and depth of knowledge.

https://www.youtube.com/watch?v=t705r8ICkRw

⬐ mcbishop

This interview is quite the exclamation mark on the statement: Elon Musk is an engineer.

⬐ mycall

Excellent interview series

⬐ simondotau

My favourite moment is from part one (linked above) about the Tesla production line. Discussion of Elon's principles of simplification (as applied to SpaceX) begins at 13:30 and those principles lead to three contextually relevant anecdotes about Tesla between 22:00—28:20.

I agree that validation is important (e.g. avoid optimising things which shouldn't exist https://youtu.be/t705r8ICkRw?t=1022 ).

It's all about bang-for-buck: a formal proof can give me lots of confidence about some aspect of a system, but might take a long time and have a maintenance burden; if that time and maintenance could give me more confidence when used to write masses of tests instead, then it's probably better to go with the tests.

My go-to approach is property-based testing, since that can exercise edge-cases more effectively than hand-picked examples, whilst taking about the same amount of effort to write.

A couple of things in favour of "proper" formal methods (in niche circumstances):

- Sometimes the effort calculation tips in favour of a formal methods; e.g. reliably testing concurrent, distributed systems is notoriously hard, and might require even more effort than doing some proofs instead.

- Proof objects themselves, and the act of writing them, can be useful for understanding a domain or system, and why certain things are/aren't true. This doesn't apply so much for "yes/no" solvers, like SAT/SMT/resolution/etc. but is often the case for "interactive" systems like Coq/Agda.

Everyday Astronaut's walk and talk with Elon Musk where he attempts to explain his management process, starts at 13:30 of part 1: https://www.youtube.com/watch?v=t705r8ICkRw&t=13m30s

Here's a rough summary:

1. Make requirements less dumb.

2. Delete part or process steps. If you're not occasionally adding things back (he says 10%) (ideally in improved versions), you're not removing things often enough.

3. Simplify, optimize, solve. Everyone's trained to jump to this because the educational process requires you to answer a question as posed, when often the question is dumb and shouldn't be dealt with as-stated.

4. Accelerate process

5. Automate

Those tend to blur together at the edges. I'm sure if he formalized this and wrote it down for mass consumption it'd be presented differently, but it's his current mental model.

Process testing - remove unnecessary in-process testing after production line debugging is done. Obviously there are nuances, he's not saying to do no in-process testing, but rather to remove testing which was intended to reveal information once that's already been collected and addressed. He cautions about false positives from in-process testing, and notes most testing can be done end of line with acceptable results.

Finally, it's important to understand the context. The part about part/step deletion in particular, when things get added back 10% of the time, is not appropriate for all development processes. That would have to be adjusted a the specific product and market objective.

⬐ adolph

That was so good I transcribed/transliterated it for my team:

Elon Musk’s Five Steps to Optimize:

- Question the Requirements - Delete Parts or Processes - Simplify - Accelerate - Automate

Quotes:

You want everyone to be a chief engineer. So, “everyone is a chief engineer” means that people need to understand the system at a high level to know when they are making a bad optimization. Like when we put immense effort into reducing engine mass but hardly any effort into reducing propellant residuals.

All designs are wrong, it’s just a matter of how wrong.

Transcript of Five Steps description:

What I’m trying to have us all implement rigorously is the five step process.

First, make your requirements less dumb. Your requirements are definitely dumb. It does not matter who gave them to you. It’s particularly dangerous if a smart person gave you the requirements because you might not question them enough.

Then, try very hard to delete a part or process. This is actually very important. If you’re not occasionally adding things back in, you’re not deleting enough. The bias tends to be very strongly towards “Let’s add this part or process step in case we need it.” But you can basically make “in case” arguments for so many things. And for a rocket that’s trying to achieve, trying to be the first fully reusable rocket…you have to run at tight margins because if you don’t run tight margins you get nothing to orbit.

Whatever requirement or constraint you have must come with a name, not a department. Because you can’t ask the department, you have to ask a person. And that person who it putting forward the requirement or constraint must take responsibility for that requirement. Otherwise you can have a requirement that basically an intern two years ago randomly came up with off the cuff and they aren’t even at the company any more . . . these things are way more silly than you’d think.

If you’re not adding things back in at least ten percent of the time you’re clearly not deleting enough.

Only at the third step you simplify or optimize. The third step, not the first step. The reason it’s the third step is because its very common, its possibly the most common error of a smart engineer is to optimize something that should not exist. And you say why would people do that? Well, everyone’s been trained in high school and college that you’ve got to answer the question, convergent logic. So you can’t tell a professor “your question is dumb.” You will get a bad grade. You have to answer the question. So everyones basically without knowing it they’ve got a mental straightjacket on. They will work on optimizing the thing that should simply not exist.

Finally you get to step four which is accelerate cycle time. You are moving too slowly, go faster. But don’t go faster until you’ve worked on the other three things first. If you’re digging your grave, don’t dig it faster, stop digging your grave.

The final step is automate.

⬐ daveguy

What does "make your requirements less dumb" mean? Make them more applicable? Make them more precise? Something completely different? Question them, but what is the flaw of "dumb" in a requirement?

⬐ colordrops

Either there isn't a precise goal so the requirements are scattered, or there is a precise goal and the requirements don't match the goal.

⬐ maroonblazer

I had the same question.

Does it refer to eliminating or deprioritizing requirements that are not core to the product? Removing "gold-plating" requirements? Or does it mean converting solutions/implementations disguised as requirements into actual requirements? Something I see a lot of in my role.

I see value in both exercises.

⬐ actinium226

I'll take a stab at answering this by paraphrasing Musk.

He said this 5 step process was inspired by the time they were trying to optimize part of the Model 3 production. There was a segment that was causing a lot of problems - the installation of some sort of foam or padding on the battery. I guess the glue wasn't adhering well between the metal of the car and the foam, or whatever I dunno, problems.

They tried all sorts of things to optimize the process, until eventually Musk asked, "what's this for?" and the Noise and Vibration team told him they weren't sure, probably for fire safety. The Battery team (the one that would institute a fire-related requirement) told him they weren't sure, it was probably for noise and vibration.

They built a car without the foam, tested it for noise and vibration against a car with the foam, couldn't tell the difference, and ultimately removed the part entirely.

So this story touches all 5 of the points, but on the point of "make the requirement less dumb" it means that whoever got the requirement in the first place should have asked "what is this for?" and should have tried to find another way to meet the requirement without needing foam (and in doing so would have found out that the requirement was BS in the first place).

⬐ adolph

I think the parent comment's story describes the concept really well. The below clip from "Pentagon Wars" describes some of the organizational dynamics that are another way to arrive at dumb requirements.

Pentagon Wars - Bradley Fighting Vehicle Evolution: https://www.youtube.com/watch?v=aXQ2lO3ieBA

⬐ RandomLensman

Also note the things mentioned by him that happen kind of before/in parallel, i.e. costing and planning.

Building high performance organizations really often is about the "mundane" stuff:

- Planning processes

- Assigning responsibility and putting incentives in place

- Clear objectives with good KPIs

I feel the same way, and I'm reminded of a quote from Elon Musk recently when discussing Starship's Raptor engine:

"It's 10-100x more difficult to design the production system than the engine itself"

So what we really need is to invest and innovate in the production of complex telescopes/probes/etc such that it doesn't cost $10 Billion and 20 years to build, and the loss of one wouldn't hinder the overall production.

https://youtu.be/t705r8ICkRw?t=238

⬐ stopping

The cost doesn't come from a lack of production efficiency. A significant fraction of the cost is for testing/validation of the spacecraft, which must be done for every unit you produce. Simulating operating conditions on Earth is extremely expensive and time consuming.

⬐ gpm

If you can produce them cheaply but not test them cheaply, the solution is really simple, just don't test them. Send them to space, see what happens.

It's not like space telescopes need any of the valuable orbit space, and even if they did it's easy to put thrusters on board that can de-orbit them or move them to a graveyard orbit (depending on their operational orbit) in the case of failure.

⬐ stopping

I addressed this in another comment: https://news.ycombinator.com/item?id=28452001

To summarize, you have to have the ability to correct statistical aberrations after manufacture via extensive validation and testing, as opposed to design aberrations which can largely be eliminated through destructive iteration. Without the former process, you'll launch a million telescopes and not a single one will work.

⬐ aero-glide2

When you build 20 scopes you can use the fail fast approach to better iterate on the next versions. It's okay if the first few fail.

⬐ petschge

Except after the second failure you have a REALLY hard time to get funding out of politicians...

⬐ loonster

Its about setting proper expectations. Its not a failure if its expected. Get the politicians to commit to all 20 launches in advanced.

⬐ rizky05

How do you know it only needs 20 launches ?

⬐ holler

It's the same reason SpaceX is eating Boeing's lunch with their space program. One uses an iterative build process, and the other uses the more historical waterfall method. That's not to say waterfall didn't or doesn't have merit here, just that it seems like technology has advanced to a place where hopefully in the future we can build these things much faster and cheaper, such that the stakes aren't so high for a single unit.

⬐ stopping

I don't disagree with your point re: Boeing v SpaceX, but precision scientific instrumentation is a very, very different domain where "move fast and break things" cannot be naively applied. The sensitivity to failure is so ridiculously high that you need 6 to 7 sigmas of reliability for each of ten thousand critical components in order to even have just a 90% chance of mission success. Something as simple as an instrument bring slightly out of tolerance, faulty rad-hardening on a CPU, the sun shield having a minor tear, are all enough to completely jeopardize the entire mission. As a fully integrated system, JWST is at least an order of magnitude more complex than Starship, whose main design goals (launch, orbit, land, don't explode) tend to favor a more iterative approach.

⬐ hfkgkrnememm

Maybe it's cheaper to launch 10 not so perfect telescopes and have 9 fail and 1 work.

Kind of like in computing, where we moved from a single ultra reliable mainframe to tens of commodity hardware.

⬐ mnl

No, it's actually more expensive because you don't get anything done. Precision equipment either works or it doesn't.

Commercial software development practices might be relevant where you can fix things overnight, but this is hard science instrumentation made of atoms, you ship things that must work as expected because the whole point of all this is making better measurements as the field evolves and there's a scientific case for doing it. It is not about selling production versions of the experiment to potential customers.

⬐ kortex

Then we need to design precision equipment differently. More redundant parts. More wiggle room. More fuzz testing. Find out where precision actually matters, and not just overbuild everything. And iterate more.

⬐ mnl

What do you think people do in tests? Why on this site folks without background feel compelled to tell professionals in whatever fields how to improve everything with generic buzzwords?

So you think that people designing a sensor have no idea about systematics and the tolerances they need. Well, they do, mainly because you start worrying about that when you're a physics freshman writing lab reports with uncertainties to calculate.

Another commenter suggested using AI. Great, now we can implement whatever biases and data dredging with training data sets. Of course obfuscated non linear regression models somehow must generate the same ontological results than actual measurements of stuff we still don't know about because... well, I have no idea about why the Universe would do that for us.

Then there's a complaint about this not being nuclear-powered. I'd like to know more about that, like using which RTGs and with what sources, because afaik there's not enough 238Pu and taking into account the power requirements of this thing, efficiencies, thermodynamics, time and economics this very design and mission could be the cheaper and frankly brilliant alternative.

I mean, selected, trained, smart, hardworking people in competitive environments generally know what they're doing and they don't need to imitate whatever the public thinks Musk has brought to humanity to achieve what only them have been caring about for years, often their whole careers.

Some experiments are very expensive and take longer than expected, you can attribute that to incompetence or to the fact that what they're trying to achieve is difficult, necessarily costly and it has never been done before with such finesse. I think it's appropriate to call out incompetence only when you can point out precise instances of incompetence.

⬐ AdamJacobMuller

Sure, "move fast and break things" can't be naively applied but that doesn't mean it can't be applied.

Instead of sending up 1 super precise super accurate super reliable telescope let's send up 1000 mostly POS telescopes and combine the results using AI of the 900 which actually make it to orbit and return useful images.

I'm hardly saying that's a 100% working plan, but, I think that kind of paradigm shift which is required.

⬐ bcraven

But how are you going to validate your scientific discoveries if all the images are coming through the pipeline: "POS telescope > AI upscale"?

⬐ HPsquared

The AI would lack training data.

⬐ hypertele-Xii

Sounds impossible to park 900 satellites on the same L2 Lagrange point. Hope you realize this telescope is to be positioned on an infinitesimally small and unstable point in empty space, on the "far side of Earth" from the Sun. Because that's the spot with the best view, apparently.

⬐ zardo

They don't park them right on L2, JWST will orbit L2 thousands of kilometers away from the exact point.

⬐ gfodor

Pretty sure this argument could have been levied against rocketry. For scientific telescopes most likely a frame shift would be needed to break out of this mindset that would be analogous but distinct from the one applied to rocketry.

⬐ stopping

There are two key differences between rockets and space telescopes (any space-faring probe, really) which prohibits "iterate to failure" as a development technique:

1. Forensics difficulty. You need extensive data to debug issues for complex systems. You can collect so much more data from terrestrial testing, which allows you to do the extensive forensic analysis required to achieve the required component reliability. Once you put a telescope in space, you can't inspect it anymore. A lot of the sensitive components which might fail on JWST have to be inspected to microscopic precision in order to perform adequate failure analysis.

2. Design requirements are far, far more precise. Many failures in deep space are effectively impossible to correct in later iterations due to point 1. Telemetry and sensor data is enough to debug rockets, but for JWST you would need to ship so much extra data/sensor infrastructure alongside the telescope that the whole project becomes recursively intractable.

The system complexity is so high that you absolutely must have the ability to make arbitrary system corrections because the chance of "building everything correctly the first time" is effectively zero, even with perfect hindsight! Essentially, any time you build the thing from scratch, you will always find mission-ending statistical deviations. The objective of on-ground testing is to identify and correct those specific deviations, until the whole system is within design tolerances. If you were to totally rebuild it, the next iteration will have a totally different set of statistical deviations which will need to be corrected. This process of development involves "hardening" the entire system through extensive testing, because it impossible to build a fully-hardened system to start with, even after "learning" from previous attempts.

⬐ gfodor

I’m not arguing that the frame shift is identical to rockets, just that assumption breaking probably is a way to avoid such risk laden events like this launch. For example, finding a way to create a flywheel to drive costs down, or a way to take smaller steps, or a way to incentivize more disposable missions that will build on each other but can tolerate some failures. I’m not an expert but the reason rocketry is moving forward again isn’t because of “move fast and break things” per se but from a rethinking of foundational assumptions in general about how rockets “must” be developed that led to that methodology being discovered as a useful one.

⬐ stopping

I do agree with this. Most systems have significant basis in unnecessary or outdated methodologies. I think with the JWST project (with partial hindsight), we probably could have benefited from having a "stepping stone" optical telescope after Hubble which we could have used to proof out some of the hard parts of JWST. We learn quite a lot from just running missions beginning-to-end, and extremely long cycle times sacrifice this learning opportunity. It also means that we focus less on developing extensible "platforms" in favor of one-of-a-kind systems which have somewhat less carryover knowledge for the next project. Shorter mission timelines mean that you can better leverage state-of-the-art technology, rather than being forced into a design which constrains you to decade-old technology.

⬐ marcodiego

This is not a device that will be produced in an assembly line. Efforts to improve its production would be nullified by the fact that every unit is itself a prototype.

⬐ devnulll

There's another way to think of this. Let's use Starlink as an example, which is ~1300 today growing to 40k satellites in LEO.

If each of these had a high-grade commercial camera lens pointed towards earth, you could have real-time viewing of the entire planet with just a little bit of math. With a bit more fun (and for each click of optics improvements), you could "Bullet Time" view a great number of things...

If each of these had a high-grade commercial lens pointing outwards, what could that be used to capture? At scale, it seems like doing diff's against all the images would reveal a huge number of previously unidentified asteroids and other near-earth objects.

What if you put "very good" optics on them? Or a radio antenna pointed out? Could you make earth sized optical or radio telescopes? Seems like you could.

It's increasingly looking like the assumption each one needs to drive the state of the art, and each one needs to be "space grade at 100x the cost" is a fallacy simply because launching was historically so expensive.

⬐ perihelions

Biggest issue is that the efficiency of telescopes at detecting faint point sources scales as (mirror diameter)^4th power -- strongly favors large telescopes over arrays of small ones.

With asteroids, you don't need to diff telescopes at different locations; you can just diff the image taken at one location, 15 minutes apart.

⬐ krastanov

Very frequently in this type of conversations a layman suggests "just make a constellation of low/average quality telescopes that acts like one big telescope". Presumably because they have heard of "long baseline telescopes"[1]. No, that does not work in the optical and infrared because you can not keep the signals from different telescopes coherent with one another (the technology to do that is science fiction today and it would be crazy expensive the first time realized). This is not to say there is no merit in the idea that many cheap telescopes can be useful in some other (non-visible-or-IR-spectrum) way, but the typical bombastic suggestions are simply wrong and are a disservice to an otherwise interesting idea. We will continue needing big-mirror optical and IR telescopes for quite a while if we want to continue studying the universe.

[1]: https://en.wikipedia.org/wiki/Very-long-baseline_interferome...

⬐ loonster

For the cameras facing earth, you wouldn't need fantastic resolution for it to be useful. With a resolution of 50 meters, you could track suezmax ships. With a resolution of 5 meters, you could track commercial aircraft. Not great, but it would help with a Malaysia Airlines Flight 370 like incident.

⬐ BobbyJo

I wonder about that. Hubble, after all, could have been looked at the same way from the outside, but it was essentially a proof of concept for a constellation of top secret satellites still in use today.

⬐ anonymousiam

https://www.americaspace.com/2012/06/06/top-secret-kh-11-spy...

⬐ asmithmd1

The opposite is true. Hubble is a stripped down version of a Keyhole spy satellite that was already in production. One of the major repairs (replace solar panels) was required because of long duration exposures used in astronomy that are not used in spying.

Remember that Hubble was essentially useless until heroically repaired the first time by Story Musgrave. Many subsequent repairs were required to get the best pictures from it - like the deep field, long duration picture.

⬐ AdamJacobMuller

I largely agree, even though you're being downvoted.

What we really need is to standardize as much of the telescope as possible.

Communications? Don't need to be bespoke per satellite. Guidance/Navigation? Doesn't need to be bespoke.

Ultimately only the actual optical payload really needs to be bespoke. Commonality of components and design can drive down prices and reduce time-to-launch.

⬐ Rebelgecko

The problem is that if you can't really just standardize every subsystem unless you're willing to give compromise your mission in the name of a one-sizw-fits-all approach. Presumably the people who built JWST passed on of-the-shelf busses for a reason. Imagine if you to tried to force JWST's bus onto every satellite:

people who don't care about pointing accuracy end up spending zillions of dollars so they can have fraction of an arc-second precision

satellites that are in Earth orbit have different thermal needs than one in L2

satellites that are less weight sensitive waste money on carbon fiber

an instrument that needs to be cooled down to 50K doesn't need a JWST-tier cryocooler, etc.

You could try having a range of things, e.g. small, medium, and large bus with a handful of interchangeable subsystems. But you're still asking designers to sacrifice their precious size, weight & power requirements

Here's a great complimentary nugget of wisdom Elon Musk shared on the tour he gave of Starbase recently that intersects with a closely related concept.

17:29 Possibly the most common error of a smart engineer is to optimize the thing that should not exist. And say, well, why would you do that? Well, everyone has been trained in high school and college that you gotta answer the question, convergent logic. So you can't tell a professor, "your question is dumb", or you will get a bad grade. You have to answer the question. So everyone is basically, without knowing it, they got like mental straight jacket on that is they'll work on optimizing the thing that should simply not exist. https://youtu.be/t705r8ICkRw?t=1049

⬐ bickeringyokel

> Most common error of a smart engineer is to optimize a thing that should not exist

I encounter this so often. It must be hard for some people to question a given design or requirement. Sometimes when someone asks for a solution to a problem, they don't actually want what they asked for they just want a solution to the problem.

This series is said to eventually have three parts. Part one [0] and two (this submission) were filmed at the production site "Starbase" (Boca Chica village, Texas). The third and last part is going to be at the launch site, a few kilometers/miles further to the east, right next to the beach. Part three has not been released yet.

Tim Dodd usually makes a good interview partner for Elon Musk because he is more knowledgable about rockets than most reporters and can ask way more interesting questions than they could.

If you haven't seen the first part yet, I would recommend to start with it [0] instead of jumping right into this one.

[0] Part 1: https://www.youtube.com/watch?v=t705r8ICkRw (53 minutes long!)

And unstacked !

It was surprising to see that it only took them about an hour to unstack (I didn't catch the stacking but I think I heard it was around the same amount of time ?).

I'm still very perplex about the "catching" mechanism they are designing, Musk mentioned in the Everyday Astronaut interview [1] that it was a very hard problem and I'm very curious to see if/how they can solve it.

[1] : https://www.youtube.com/watch?v=t705r8ICkRw

⬐ MayeulC

Nice video, thanks for sharing. I liked that bit on requirements:

> Whatever requirement or constraint you have, it must come with a name, not a department. 'Cause you can't ask the departments, you have to ask a person, and that person who's putting forward, the requirement or constraint must agree that, they must take responsibility for that requirement.Otherwise you could have a requirement that basically an intern two years ago randomly came up with, off the cuff and they're not even have the company anymore. But it came from the, let's say, air loads department. They're like, actually there's no one at our current department that currently agrees with that.

⬐ Fiahil

As a "consultant" I fight phantom requirements every day! I can't agree more with this method!

⬐ ipodopt

Yeah, I think the chopsticks makes sense. For the actual contact I would have thought a friction based solution (wide soft/grippy chopsticks) being easier in practice. Too bad the metal they are using is not magnetic.

⬐ eightysixfour

I'm sure you've seen this rendering and Elon's response, but I thought I would post it for other's who have not: https://twitter.com/elonmusk/status/1422603106035118085

⬐ cwizou

I had not, thanks ! This seems crazy, I can't wait to see it in action.

⬐ phendrenad2

Looking at that, surprised they didn't go for a taller tower and some kind of cable arrest system like aircraft carriers have. I'm sure they thought of it.

⬐ LeifCarrotson

I am interested as well - terrestrial civil engineering, and associated carbon steel, concrete, and rebar materials, are dirt cheap in comparison to interplanetary aerospace engineering and their alloy or composite materials.

I'd have started my design like a Delta robot or inverted Stewart platform, with a trio of sturdy towers, a catchment ring in the middle, and cables running to large electric motors with massive heatsinks (saltwater?) that could position the ring anywhere in X/Y/Z between the towers. By choosing a site with a low central valley between three natural peaks, you could get a huge amount of volume and not need as high a tower, I hear there's a site over a Puerto Rican sinkhole that recently became available - though be gentle during disassembly of the previous structure, it belongs in a museum....

Engineering is expensive, materials are cheap; always make the process window as large as possible!

⬐ SideburnsOfDoom

I've seen that rendering and it looks absolutely nuts. "catch the falling rocket with tongs". It would be amazing if it works, but for now it looks like something that sci-fi would avoid as not seeming plausible.

I can see this as being a real story.

I just finished watching this really very interesting interview with Elon as he walks around the spacex stuff https://youtu.be/t705r8ICkRw

He got vocally and visibly annoyed with the beeping noises that the machines were making. It was a useless cacophony of beeps and you could just see in his head he was thinking "How the hell do I make this better and quieter"

In a recent interview [1] Elon stated that empty booster "has the density of an empty beer can" which is a great mental aid. So it will certianly float.

But part of it will be underwater, initially the very expensive flamey part (engines). They will not "be the same" after the experience.

[1] https://www.youtube.com/watch?v=t705r8ICkRw

The recent interview of Elon by Tim Dodd is probably more informative and interesting this this could ever be.

https://www.youtube.com/watch?v=t705r8ICkRw

⬐ milansuk

I remember watching "Everyday Astronaut" live with less than 1000 other people. Today, the numbers are over 100K. He is a great guy, who is very interested in and passionate about everything around space exploration and his articles and videos go very deep, the same for answering questions live. Only if more YouTubers were like him.

Can't wait for 2nd and 3rd parts.

His notes for 1st part: https://everydayastronaut.com/starbase-tour-and-interview-wi...

⬐ tpmx

Shouldn't they have been wearing helmets when filming that segment? It was a quite busy work site with high altitude work (so a risk of dropping tools etc).

⬐ IanClarke

In the beginning of the video they were high up and I feared Elon could fall to his death and that would bring "all of human's progress to a halt". They not only work 24 hours and 7 days a week (In 12 hour shifts: 3 days and 4 days in a row, then 3/4 days off), but they also rush things, with tight deadlines, ofc. top-notch workers/engineers, but still. No fatalities while working like this (I googled it the other day) - It all looks dangerous.

⬐ tpmx

Yeah. But wearing a helmet isn't cool, so...

⬐ IanClarke

Elon's new, transplanted hair.

He is very proud of it in the video, combing through it with his hand all the time. Especially with his current new hair-style: Short on the sides, long on the back and top - a rebell, cyberpunk, cybertruck-t-shirt on.

He looks great,... today. Gone are the years where he was bullied keeping his nose in the books and -though coming from a modelling-family (his mother)- being (looking) rather nerdy/"average"... a bit sleepy (when he was younger), going from a "nerdy" looking teenager to losing his front-hair as a young man.

I envy people that can fix their appearance with a beauty surgery. Most can't and of those that try, a lot do not look so good afterwards.

⬐ IanClarke

I was surprised how Elon talked "in depth" about technology and engineering in this video with Tim, since Elon often uses relatively simple things to talk about to reach the masses, like "rockets must be fully reusable", "we need sustainable transportation (and they are electric cars)", etc. Elon likes to use simple language and paint big pictures. But he (and his team) must have evaluated thousands of interdependent complex factors and found the one solution (e.g. Methane-propelled Starship) before reaching to these simple "big pictures" (which he then likes to talk about and present to the public) and ofc. it's the right thing to do and there is nothing wrong with it.

Elon is the new Steve Jobs, informing and educating us about the future, just like Steve did it. This time with cars and space exploration, which is just "next" (Steve Jobs informed us about the iPod and the iPhone). When Steve wiped out Nokia with the iPhone, I thought that someone from Silicon Valley would do the same for cars and there are a couple of contenders now with Tesla leading. But it's a big market with a lot of companies, and instead of wiping one out, there are just many suffering now - more or less, or still doing fine.

Though there are very much details, which he is aware of and trading/evaluating... at least with his engineers and which he does in this video, which is nice, refreshing and new information (for us and from him).

I really like Tim Dodd, though he comes from a background of "I wasn't good at school and am still excited and knowledgeable about rockets today [and so can you(everyone)]", he is sometimes over-explaining simple things imho. I have also the feeling that Tim Dodd might be a sub-par interview partner for Elon in this video, but Elon manages it fine and I think he likes Tim for everything he achieved despite not having an academic background.

Just recently I found https://youtube.com/c/MarcusHouse, which fits my taste a bit better, currently.

⬐ hackeraccount

Tim Dodd reminds me of Andy Weir. They're both trying to explain things that are fairly complicated to an audience which is interested but might be starting from zero. That's actually really hard to do and then they add to the degree of difficulty by not being afraid to release the minimum viable product to that end.

For me it makes both of them fun to watch because you can see them getting better. As a novelist Weir is - in my opinion - improving his craft with each novel. Same with Dodd when it comes to the videos he's making. Both really impress me.

⬐ spikels

Came here to post that video. Awesome!

My favorite part:

Elon's 5-Step Design/Development Process:

(1) Fix "Dumb" Requirements

(2) Remove Unnecessary Parts/Processes

(3) Simplify/Optimize

(4) Speed Up Cycle Time

(5) Automate

⬐ inglor_cz

Two things I took from that interview:

* the quip about people spending a lot of time to optimize something unnecessary,

* the quip that you must be wary about recommendations made by smart people, because you will tend to trust smart people too much and they can still make mistakes.

⬐ jeffrallen

Also leadership through humbleness and vulnerability: "I myself made that mistake, I followed the process exactly backwards and wasted a lot of time."

Also leadership by storytelling. Don't say what you want, tell a story that shows people what you want, so they can envision themselves doing what you want.

⬐ gkiely

"If parts are not being added back into the design at least 10% of the time, not enough parts are being deleted."

I am struggling to understand this one.

Is he saying that you should remove parts on a routine basis, like a "remove parts/processes" meeting every month and then re-adding them back in once you find they're needed.

Or is he saying when you get a list of requirements, remove parts/processes related to those requirements and then re-add them once you find they're needed?

Or is it something else entirely?

⬐ Retric

He’s talking about physical parts and manufacturing processes. For example laser etching a part number might might not be worth it. If nobody can justify why it’s needed then why not skip it.

⬐ gkiely

Got it, thank you!

⬐ smarri

Awesome video and thanks for the list, I was just going to write that down and you saved me a job.

⬐ Ajedi32

Tim Dodd's website also has an article version of the video with a more detailed summary: https://everydayastronaut.com/starbase-tour-and-interview-wi...

> Musk overviewed his five step engineering process, which must be completed in order:

> 1. Make the requirements less dumb. The requirements are definitely dumb; it does not matter who gave them to you. He notes that it’s particularly dangerous if someone who is smart gives them the requirements, as one may not question the requirements enough. “Everyone’s wrong. No matter who you are, everyone is wrong some of the time.” He further notes that “all designs are wrong, it’s just a matter of how wrong.”

> 2. Try very hard to delete the part or process. If parts are not being added back into the design at least 10% of the time, not enough parts are being deleted. Musk noted that the bias tends to be very strongly toward “let’s add this part or process step in case we need it.” Additionally, each required part and process must come from a name, not a department, as a department cannot be asked why a requirement exists, but a person can.

> 3. Simplify and optimize the design. This is step three as the most common error of a smart engineer is to optimize something that should not exist.

> 4. Accelerate cycle time. Musk states “you’re moving too slowly, go faster! But don’t go faster until you’ve worked on the other three things first.”

> 5. Automate. An important part of this is to remove in-process testing after the problems have been diagnosed; if a product is reaching the end of a production line with a high acceptance rate, there is no need for in-process testing.

⬐ postmeta

1a. tie requirements to a person not group

⬐ __sy__

I thought this was an absolutely masterful explanation of engineering design thinking. If I may, I would also like to add to this list another thing he mentioned: "each requirement must have a name attached to it."

I believe he was referring to the first step. And "Name" was really "a person" as opposed to "a department." This was his way of saying DRI all the way down to each individual requirements, to force someone to be accountable for it. His example was that SpaceX interns previously listed requirements, that were later assigned to a department (e.g. avionics), and when prompted, no one in that department really knew why it was there.

⬐ askytb

Elon is one of the best people to learn from about the development speed, given how fast they move at SpaceX, so I thought I'd post it here.

My summary of what he talks about in his 5 steps:

1) Make the requirements less dumb. I.e. always question the requirements and always have a name of a person responsible attached to them (not the department), so you know whom to question.

2) Try very hard to delete a part or a process. If you are not adding things back 10% of the time, you are not deleting enough.

3) Simplify and optimize, but ONLY after doing the first 2 steps.

4) Accelerate cycle time, you can always go faster.

5) Automate.

All five only in that order, he talks about making the mistake of going through the steps backwards and then deleting the part.

Overall it's a great interview at the Starbase Factory, so I recommend you watch the whole thing, or at the very least the relevant ~12min portion.

Did you just watch this tour of the Starbase with Elon Musk?

He spends about 5 minutes talking about how removing things is the most essential part of a good design: https://youtu.be/t705r8ICkRw?t=833

Any source for that? He sure seems to know very fine detail of almost anything related to the product.

See for example:

Everyday Astronaut recent Musk interview: https://www.youtube.com/watch?v=t705r8ICkRw

Munro interview with Musk: https://www.youtube.com/watch?v=YAtLTLiqNwg

There are also other examples where Musk talks about things it's hard to imagine of him even being aware of if he's not building anything.

⬐ int_19h

You're right, I should be fair: Musk the engineer is building something, but alongside many other people, and his contribution there is not dominant. Musk the entrepreneur isn't building anything, though, and that's what the man mostly is.

⬐ polytely

It's so cool to get a tour like this, fascinating stuff.

The part at 12:47 seems to be some really valuable knowledge, when he says: if you are not occasionally adding things back in, it means that you are not eliminating stuff aggressively enough.

Smarter Everyday did a tour of the ULA facilities [0], which is also worth checking out if you like this kind of thing.

[0]: https://www.youtube.com/watch?v=o0fG_lnVhHw

⬐ nickik

I'm only 10 minutes in and this is already amazing.

They will defiantly need batteries optimized for the booster. They are basically caring a bunch of car batteries around.

⬐ guardiangod

It's good to see Elon Musk not to take personal credit on anything in the video. It shows that he is just a part of the team, and not the person who 'created' the rockets.

As a side note, aren't hard hats required in the site?

⬐ wmf

If Elon hates beeping forklifts and yellow safety markings I guess hard hats are also out.

⬐ guardiangod

From the video and https://everydayastronaut.com/starbase-tour-and-interview-wi...

Musk overviewed his five step engineering process, which must be completed in order:

1. Make the requirements less dumb. The requirements are definitely dumb; it does not matter who gave them to you. He notes that it’s particularly dangerous if someone who is smart gives them the requirements, as one may not question the requirements enough. “Everyone’s wrong. No matter who you are, everyone is wrong some of the time.” He further notes that “all designs are wrong, it’s just a matter of how wrong.”

2. Try very hard to delete the part or process. If parts are not being added back into the design at least 10% of the time, not enough parts are being deleted. Musk noted that the bias tends to be very strongly toward “let’s add this part or process step in case we need it.” Additionally, each required part and process must come from a name, not a department, as a department cannot be asked why a requirement exists, but a person can.

3. Simplify and optimize the design. This is step three as the most common error of a smart engineer is to optimize something that should not exist.

4. Accelerate cycle time. Musk states “you’re moving too slowly, go faster! But don’t go faster until you’ve worked on the other three things first.”

5. Automate. An important part of this is to remove in-process testing after the problems have been diagnosed; if a product is reaching the end of a production line with a high acceptance rate, there is no need for in-process testing.

Additionally, Musk restated that he believes everyone should be a chief engineer. Engineers need to understand the system at a high level to understand when they are making a bad optimization. As an example, Musk noted that an order of magnitude more time has been spent reducing engine mass than reducing residual propellant, despite both being equally as important.

⬐ soheil

This is what I came to hackernews to find, I heard him mention it in the video I think for the first time.

⬐ _antix

I love how Elon just geeks out about engineering. So many sessions for software development as well.