protastus

Follow the money.

The boutique US universities are private institutions sitting on unbelievably large endowments. They're strange hybrids between school, hedge fund and country club.

Admission is a committee driven process to decide who can join the very exclusive club.

https://en.m.wikipedia.org/wiki/List_of_colleges_and_univers...

The issues raised and the woke bias can be simultaneously true, making this a false dichotomy.

VP and GM are not middle manager titles at Boeing.

It makes sense. Engineers who get MBAs (or any second degree) usually do it early in their careers, so it's not like they have deep experiences to anchor them after they switch career path.

You claimed it wasn't clear how Qualcomm could help. I pointed out how they could help. There's precedent too, with ARM supporting panfrost.

QC is not a low level employee lacking agency over their job description. They're a business that is free to set their strategy in response to risk and competition.

The less that QC upstreams, the more risk they incur, especially with tightening security scrutiny (e.g. 2021 Cyber EO requirements).

QC can also get disrupted by Mediatek doing a better job with Linux.

There are so many ways they could help. For GPU, they could support freedreno (both the kernel and Mesa parts) instead of KGSL and the closed source Adreno libraries.

A great start would be to make the user space Adreno GLES libraries talk to the freedreno DRM backend. So one could use the QC proprietary GLES libs with a mainline kernel without rebuilding with KGSL.

There's a similar story here for every subsystem driver, where they insist on a brittle proprietary solutions (one branch for every chip!) that can't be upstreamed.

100% agree with you.

Qualcomm historically wrote awful OSS code that couldn't be upstreamed because it barely worked on a subset of their own SoCs, and broke basic function for everything else. Typically featuring incredibly invasive architectural changes that patched far too many layers of the OS with unmaintainable code.

QC then converted bad code into a business model, nickel and diming OEMs who needed bugs fixed to ship their products.

The idea that QC SoCs now have good upstream support is bizarre to me. At best, it means a few high volume SoCs can boot to console on reference designs. Forget about peripheral support for things anyone would take for granted, like audio and GPU.

I love ramen, it's one of my favorite foods.

There seems to be a pattern of ramen shop owners/operators who are obsessed about every aspect of their product, work insane hours and refuse to relinquish control because nobody else could meet their standards.

I understand that no employee could have the same obsession and commitment. But I wish they'd hire some staff anyway.

Tesla has the valuation of a tech company. Traditional automakers have the valuation of heavy industries.

Traditional automaker execs want Big Tech valuations. So they've loaded up their products with tech, in areas they have no track record of delivering, and often without any customer demand. Starting with EVs.

Result: overpriced cars that are substantially less reliable.

This sloppy delivery will further separate the wheat from the chaff in this space. We already see a narrative from certain automakers that EVs aren't competitive. They should be more specific: their EVs aren't competitive.

I'm always happy to see views of human exceptionalism getting torn down.

Animals show behaviors that are very close to human, often identical modulo anatomical differences. Animals lack a language that we can understand, and therefore can't defend themselves against claims that their behavior doesn't hold up to arbitrary standards of emotion, intelligence, etc.

All sentient beings deserve to be treated with respect and dignity. The vast majority of humanity is not on logical or ethical solid ground here, and remains in denial.

I think most of us would love to preserve the rainforest, but lack the tools to do it. We are therefore working on parts of the problem that are still very real, very consequential, and that we can control.

Solving climate change requires switching the infrastructure of humanity's heaviest industries, that produce at enormous scale, and are the product of centuries of optimization.

Doing this without a heavy green premium requires technology and innovation, allowing the replacement solution to be cost competitive. Otherwise, the lack of economic competitiveness will prevent new methods from replacing the old at any reasonable scale. We've already seen that policy alone (i.e., taxes) can't make it happen across all the sectors at the necessary speed, due to being vastly unpopular.

> Theres no scenario I can think of where it actually makes sense to use 5000 units of energy to clean up the CO2 generated by the production of 100 units of energy

This line of thinking comes from the fossil fuel industry. Nobody in clean tech advocates for this.

The problem is that even if we get to net zero by 2050, there will be too much carbon in the atmosphere. Global warming is actually suppressed by aerosols emitted by coal and other fossil fuels, and is going to get much worse even if we meet energy transition commitments. Further, there are industries that we don't know how to decarbonize (e.g., aviation, sustaining need for things like plastics).

Carbon capture (point source and direct air, whether for storage or for reuse) are logically sound. They are extremely expensive (thermodynamically, resource, $) and basically nobody likes them, but humanity lacks better solutions.

Inspecting planes does require a lot of skill and knowledge. But one doesn't need deep expertise to detect loose bolts.

My experience is that when quality breaks down, the effects are visible at multiple scales and multiple places. Who knows what else is non-compliant in the 737 MAX assembly line (surely this is why the FAA announced an audit).

Loose bolts happen because multiple layers of protection were bypassed. https://en.wikipedia.org/wiki/Swiss_cheese_model

Solar is super cheap, making it a very effective complement to the matrix even though it's not enough by itself, for reasons you mentioned.

Every kWh from solar is one kWh less from fossil fuels. It seems inevitable that electricity becomes free when there's sun, because of solar.

I think my thesis is non-controversial and entirely obvious. So I suspect we diverge on definitions rather than have a factual disagreement.

Tesla is big tech (today). SpaceX isn't only because it is niche, and hasn't gone public. Both are 20 years old, and have for the last two decades been the most exciting companies in their respective segments, by far. This allows them to compete very favorably.

GM is boring, slow and makes many public mistakes. GM filed for bankruptcy in 2009. It has to pay a significant premium to hire equivalent talent and offers significant career risk, especially to young employees.

Or maybe he's alive and well under a new identity.

In the US, tech sucks all the oxygen out of the room (no pun intended).

When so much top talent is locked up by big tech, who's left to design planes at Boeing, write software for GM, run for office?

Yes, and allegedly the issue was root caused and fixed. The cost to investigate and recall has been sunk.

Are they time travelers from the late 20th century?

This attitude was very prevalent in the 1990's. There's no excuse for being ignorant about software in 2023 though.

There's something missing from this narrative that GM is not disclosing. It's very rare for a manufacturer to pause sales of a product that works well, is well received and in demand. GM has hyped their EV strategy and the Bolt is the one thing that's real and good.

Automakers redesign cars all the time without pausing sales. Automakers work on the new design in parallel, then switch over and call it a new model year.

Maybe a supplier raised prices and GM's gross margin tanked, or maybe they were always losing money on every car made and they need to stop the bleeding.

More likely and easier scenario: instead of hijacking, brick the cars.

Cars have become computers with a different form factor. I'm not confident that automakers can keep their SW/FW stack up to date and secure, given their complex supply chains, scrambling to build products that are new to them, while competing with big tech to retain SW engineering talent.

Rudely stated yet true. I've used Firefox as my main browser for as long as it has existed (and Netscape before). It's very concerning to see it slip into irrelevance, putting more users every year under the control of big tech.

Yup. The bar keeps going up.

First everyone started driving SUVs and pickup trucks. Then the SUVs and trucks got bigger and heavier until they could barely fit in a parking spot or highway lane.

So obviously the next step was to make a truck out of steel with sharp edges, and have an enormously heavy battery for maximum destructive potential.

America needs the CyberSUV.

This is from Tesla, the U.S. company which has been shipping a half baked safety critical feature under the name Autopilot/Full Self Driving for years, with no government pushback.

America is basically the wild west. Whether you're a pedestrian or in another car, you're on your own.

Preserve your reputation without getting sued.

> Does she see herself as more trustworthy? She can't even be bothered to give an excuse for the firing.

I agree that outsiders can't weigh in on this topic without statements from the board.

But as the (former) board member of a private organization, she's not accountable to the public. Making statements about why she fired Altman can carry substantial legal risk.

Also:

- Apollo simple with many redundancies, design for failure, culture of risk management (after Apollo 1)

- Artemis very complex, many unknowns, unanswered questions, poor communication

Does NASA and their contractors have any true believers in leadership positions?

The complexity and safety concerns have been pointed out multiple times. I stopped paying attention because this program appears to be on bureaucratic autopilot. Government contractors deliver parts, without control or accountability about the performance of the entire system. The complexity indicates that Artemis will get cancelled once it fails to achieve goals. I hope nobody dies.

The top engineering talent of the U.S. is not in this program. Poor engineering leadership is inevitable. Artemis looks like a jobs program.

If we're discussing garbage content from the NY Times, then their criticism of Robert Goddard tops the list for me

https://www.forbes.com/sites/kionasmith/2018/07/19/the-corre...

Not only that, the State Department revoked his only citizenship and made him stateless. This is a huge deal and it's crazy this is possible (legally and bureaucratically) in the United States.

The article makes it sound like an anonymous member of the bureaucracy checked a box in a form and now this man is a ghost. No legal proceedings, no transparency, no accountability.