Makani’s time as an “other bet” comes to an end
medium.com
medium.com
Bezos is inspiring, maybe thats about it for me.
(Not an Amazon.com employee or ex-employee. I have no inside information.) I was excited to learn about the culture at Amazon.com but I'm personally very disappointed (perhaps because of the high bar I imagined). The story of buying a door and using as a desk probably gets repeated a lot but that and the touted base salary cap gets lost in the sky high executive compensation.
Salary.com https://archive.md/oIQ67 pegs both Jeffrey A. Wilke, CEO Worldwide Consumer and Andrew R. Jassy, CEO Amazon Web Services at over seventeen million dollars per year and Jeffrey M. Blackburn, SVP, Business Development at over ten million dollars per year. Even Jeff Bezos took in over a million and half dollars. I don't find it inspiring at all. I would find it inspiring if the leaders made closer to the salary cap (definitely under a million dollars a year).
That said, my inspiration is independent of their compensation package
You can find engineers in faang companies making over a million, so that seems like a very restrictivr compensation package you are suggesting.
The context of the "Developers, developers, developers" thing was that MS was fighting against open source to try to attract developers to the Windows platform instead.
Visual Studio Code, Github etc only came after Ballmer left.
[1] https://www.theregister.co.uk/2013/08/24/top_10_steve_ballme...
I like Apple and Steve Jobs, it's not the same without him but in the age where we are building the capacity to colonize another planet that far exceeds "one small step for man, and one giant leap for mankind", it's baffling to see people inspired by the frugality of one's personal income.
Some of the greatest achievements in the history of humankind are on the verge of happening and most people on HN can't see past their neighbor's pocketbook.
Or are we just hoping Bezos takes us all in his rocketships?
Prediction: all "other bets" not related to Ads, Search, and Cloud will be gone and forgotten within 5 years. Deepmind will probably remain, but will refocus on stuff that has more than just PR value.
Some may claim that. But there is little support for it besides the fact that they simply structured it that way for strategic purposes and possibly to try to de-emphasize their search dominance for the monopoly hawks.
The 'Google' portion of Alphabet is the only net income producer and almost the only revenue producer. All other Alphabet business are big time money losers right now.
Google doesn't want people/regulators to think they're a monopolistic search engine, so they build that phony shell company to host a bunch of nonsense sideshows to divert attention.
Opinion my own, not Google’s.
Another, "Project Malta," which aims to store energy in molten salt, was spun out into an independent company. https://www.maltainc.com/our-story
Edit: Looks like https://www.sciencedirect.com/science/article/abs/pii/S17505... and https://www.sciencedirect.com/science/article/abs/pii/S17505...
But the thing is, schools can’t use brushless motors by themselves. They need power supplies and motor controllers too. So when my team was throwing away fifty 600 watt power supplies, I wanted those too.
The whole time, I was not just asking for one item from the scrap bin, I was asking for an approval process. We threw away perhaps tens of thousands of dollars of valuable off the shelf parts a year. I couldn’t be talking to the head of X every time I needed approval. But management in my team didn’t care. Instead of responding to my request, they complained about my methods, saying I need to go through “proper channels”. I drafted a document describing a process for saving items from scrap that had documentation and oversight without requiring much bandwidth from the higher ups. It went nowhere.
So Google X still scraps thousands of dollars of stuff on the regular. Many people inside X want to help donate that stuff to schools. But the management on my team didn’t care one bit. I guess when you’re paid a million bucks a year, you just can’t relate to the local school kids who can’t afford motors for their robot. But I was once that kid, and to see these people dismiss what help we could give them was too much for me.
It’s still upsetting. Google X talks about changing the world and moonshots but you know what would change the world? Giving valuable supplies to local schools. I offered to do all the work on my own time, but they couldn’t be bothered to give me the time of day.
As a contrast, the team I was on (Making and Science) purchased an enormous number of safety glasses, which we gave away at Maker Faire. We often gave whole boxes to teachers who asked (like, 144 glasses). You see them all over the place, and the teachers were always thrilled to get them.
But throwing away valuable stuff is still waste. If you can help someone a great deal for next to nothing, that’s worth doing. Forgoing that help and then making donations elsewhere is nice I guess, but it’s still wasteful.
Last time I've put them on was testing our open-hardware pico-satellite deployer.
Thanks for saving my eyes.
I think those examples show that R&D without an immediate return can be a rational decision for a corporation.
Another way to think about it is that Google really should be doing long-term research, because only finding and monetizing a major breakthrough is going to make it so that they don't have all their eggs in one basket. A basket that will, like all baskets, eventually break.
A third way is that as long as Google is extracting monopoly rents, they should probably put some of that money into government-like things that could have broad societal benefit. Because if they keep trying to maximize quarterly profit, they'll burn up all the societal goodwill that insulates them from real regulation.
As for ploughing investment to keep growing at the cost of profitability, that's every VCs formula and very well understood by the MBA crowd.
I'll note that the US had much higher economic growth during a period where large companies all were, by modern standards, overinvesting in R&D. It could be that we're just much smarter now. But it also could be that since CEO tenure has fallen nearly in half in the last 50 years, execs have stopped investing in things that they won't personally be around to profit from.
Makani was funded for 14 years! 14 years!! How many companies are willing to fund speculative research programmes in things unrelated to their core business for so many years? None, that I'm aware of.
Truth is Makani should have been shut down years ago. Government or corporate has nothing to do with it (that is, the same assessment would apply if it was funded by the government too). Their idea clearly wasn't competitive with the by now highly optimised wind turbine industry. It's not obvious why it ever was expected to be so.
I definitely agree that Google should prune their portfolio using rigorous standards. I'm just saying that if one judges the whole effort by typical established-business standards, you're going to get typical results. Some things need more time to prove viability. Since Google a) will be around a long time, and b) is currently dependent on a single cash cow, they can (and should!) take long-term risks with long-term payoffs.
> It's not obvious why it ever was expected to be so.
This is a very bad way to look at long-term research. Or any novel venture, really. You want to be a pundit with a great track record on startups? Just say each one will fail. You'll be right 90% of the time with very little effort.
But it's even easier to wait until something fails and say, "It's not obvious why it was ever expected to work." Of course! If it were obvious that something was going to work, it wouldn't need to be done as a startup. And once it has actually failed, hindsight bias lets us paint it as an inevitable failure. E.g., if SpaceX had had less money and a couple more explosions, it could easily have gone under, and then all of its detractors could have done the "we told you it would never work" routine.
This is not a great way to think of Amazon. Amazon is basically a VC that only funds internal "start-ups" and "unicorns." You still go to Bezos and the leadership team with a business plan, financial model, path to profitability, market fit, differentiation, competitors, TAM/TAM growth, and "is it a land grab?"
They do not fund "interesting ideas for the sake of interesting ideas" like the Google X projects. Even the most secretive Amazon projects tend to tie pretty closely to their existing businesses and strategies when they are revealed.
Also, not sure if this is still true, but Amazon was the #1 largest hirer of MBAs from top programs for a couple years.
15 years ago Bezos could have retired rich and turned it over to a standard CEO, who would have milked their then-excellent position commanding a large and growing share of ecommerce. And he would have been applauded by a lot of analysts, etc. But instead he's put titanic sums of money into making what are, at least on paper, pretty modest gains in terms of purchasing friction and latency. Is 2 days really that much better than 3-7 days? Is next-day and same-day really better still? The customers' answer is "fuck yes!" But it wasn't clear up front, at least to a lot of investors and commentators, that it was really worth tens of billions in capex.
In both cases, what I'm talking about here is investments that go beyond the time horizon of most companies. If you measure them by the average company (which has a much shorter exec and CEO tenure), of course they'll look bad.
Google should focus on something else than ads.
also: They are wasting resources with their side projects.
No one see the irony in that. I like their wacky projects. Obviously they to profit off of them at some point. But they are also kind of ground breaking if they actually turn out to work.
Alphabet also has the luxury problem of having discovered the most profitable business in the history of human kind. Now they struggle culturally to find something else to grow. They are user to basically make money for free and to so at insane scales.
Cloud is getting there but it‘s still nowhere near as easy. That‘s where Amazon in particular is very much in the advantage. They are already dominating the lowest margin business ever (online retail at scale). And they have a good foothold in cloud. Now they are getting into ads more and more. Much more comfortable direction to go.
In that sense the moonshots might make more business sense as well.
If Waymo works out globally. Add half a trillion to the valuation.
Wing: This could help them compete with Amazon.
GV/CapitalG: Very successful.
Sidewalk labs: could one day power cities.
Loon: not sure. Their better bet seems the heavy spacex investment with Starlink now. Though there are some pilots.
Verily: The health sector is fucked with leeches and inefficiencies. Especially entrenched players like epic who only want to protect their cozy position. Big opportunities there.
Calico: who knows. In a decade they could discover something insane. No idea.
Deepmind/brain: this is it. If only one project can succeed it should be this one in their view. If they succeed with AGI/ASI anything else doesn‘t matter. This is of course a big if and likely decades away.
The only thing they‘re currently not funding is fusion research. But that might be too long term even for them.
I'm assuming "the deepwater scenario" means tethered kites supplying power back to land somehow.
Transmission is always the tricky part, and I'd imagine it is even more so in a deepwater scenario where laying cables is even more expensive.
The system doesn't even need to be anchored.
Keep in mind that the nacelle of a regular windmill is already quite a complex beast, it has one degree of freedom less and can adopt to gusty winds by simply changing its pitch, there is no need to bring the whole contraption back down to the ground before the wind fails to keep it aloft. The only way you can keep the Makani kite aloft if the wind drops rapidly is by controlled descent and that in turn you can do only about as long as your cables are. Typical generators for 'serious' windmills weigh on the order of 30-100 tons, that's roughly the take off weight of a 737-800 on the high end.
Props for actually building it though, that part - even at the limited scale of the prototypes - is quite an achievement.
Most of all, investors wanted clarity on spending, and Google reacted with the restructuring.
Let's take a moment to fairly summarize what's happened: extremely risky ideas that no sane investor would fund, received disproportionate funding, and as far as I can tell, they all failed even after a decade+ of not expecting profitability. There's no reason to pin this on Sundar - the ideas failed on merit
Also, speaking honestly, he's been rather generous with many projects (Chrome OS and related hardware, Stadia, even horribly performing pixel phone sales), despite decades of losing billions for Google.
At least Microsoft knew when to cut their losses with Windows phone
(Smaller projects, because Waymo certainly wouldn't fit that framework. But Waymo is sitting right at the ad business core: "ok car, take me to an Italian restaurant" would be 100% directory ad business)
But it's also a terrifying controls problem, and suffers from the same drawback as all the "flying car" startups: What goes up, must come down. Traditional wind turbines are a bit more predictable. The offshore flight test in August ended with a crash and the loss of the aircraft. I have to assume that no net power was generated during this test, or they would have mentioned it in the press release?
https://medium.com/makani-blog/makanis-airborne-wind-power-s...
Makani's device is able to briefly generate power for a fraction of each crosswind loop it flies, but it has to draw power from the grid to push itself the rest of the way around the loop. This is neither useful as an energy source, nor cost competitive with a traditional wind turbine.
For offshore applications, there are much simpler competing projects to put traditional wind turbines on spar buoys or semi-submersible platforms, e.g. http://www.principlepowerinc.com/en/windfloat
The easiest way to compare Makani to stationary wind installations out there generating power today is to look at the effective cross section of the rotors and to realize that a moving rotor consumes exactly the same amount of power in lift that it uses going forward. So that will cancel out across the loop leaving you with a bunch of small rotors vs a much larger rotor sweeping an enormous surface. And launching it is one thing, bringing it back in in a storm is a completely different proposition.
A 12MW Makani installation would use much more land and would be much more fragile than a comparable regular HAWT.
A lot of interesting ideas get shot down quickly, which is not necessarily a bad thing.. but I'm not sure what the false positive vs. false negative rate is in this process. But for those ideas that develop, it seems like perfectly solid technologies that could form the premise of an interesting startup don't get out the door because they don't meet an internal bar of "worthiness". Maybe let the market and external funders decide that?
Their internal process is not intrinsically good or bad, but what's unfortunate is that they don't really publish. I know they're not looking to be Bell Labs, but if they think a lot of these ideas aren't going to pan out, but many person-years of research effort goes in, they really should share the learnings and lessons as opposed to locking it up internally. When these projects don't pass their internal bar for being commercially viable (which seems rather disconnected from what that bar would be if the idea/ tech stood on its own outside), it's basically like academic research. Truly is a shame that a lot of smart people working on interesting non-commercializable ideas don't get to share that. Others might pick up on it, if not now.. maybe a decade or two down the road.
In general, people in X get paid too much to make it "worth it" to innovate. Obviously there are exceptions, but I'd say there is a perverse incentive to string the funders (Google) along as long as possible. Line engineers are making $250K+ / year, managers are making $500K / year.
Innovation is kind of a "swing big and go home if you miss". But there's no incentive to "go home" if you have a comfortable lifestyle. You want to keep going and you can convince the management of that because you inherently know more about the tech than they do. It's an information assymetry problem.
I see a similar effect in non-X moonshots funded by Larry Page.
https://www.forbes.com/sites/jeremybogaisky/2019/12/01/insid...
Basically there's an incentive for "demo-ware" and not real innovation.
I'm glad I worked at Google rather than say Tesla, because I was doing "non-moonshot" engineering work. Tesla is notoriously chaotic. I think Chris Lattner said Tesla had the highest turnover rate he'd ever seen.
But I actually think the idea of firing a lot of people makes sense if you're innovating. It should be chaotic and messy because the whole goal is inherently risky.
I used to work in video games and they also fired people all the time and paid them like crap. I made 3-4x at Google what I did in the video game industry. But honestly it makes sense because of the financial profile of the industry vs. Google's industry. Put bluntly: you're filtering for people who really love it and are doing something irrational to innovate hard tech or create a blockbuster game.
I'm sure a lot of people will bristle at this from the employee perspective, but having worked in both environments for many years, I can see how the compensation changes people's behavior. You might not realize it but it does.
The irrational thing being working way below the market rate for someone with their skills/talents, right?
When the employees no longer believe in the goal, they can quit, "cash in" and go work a more traditional job. But X removes that natural dynamic, hence the questionable results being achieved IMO.
Are they shutting down? Are they getting acquired by Shell? Something else?
>Shell is exploring options to continue developing Makani’s technology.
Vague enough to be shopping around perhaps
so from today Makani’s time at Alphabet is coming to an end. This doesn’t mean the end of the road for the technology Makani developed, but it does mean that Makani will no longer be an Alphabet company. Shell is exploring options to continue developing Makani’s technology.
Can you make electricity with kites? Maybe. Probably. Should you? Probably not.
If you look at the Goole X website (https://x.company/), the problem is stated thusly: "How can kites be used to generate electricity in unexpected places?"
Any sensible person's first reaction to that should be: Why would you want to do that in the first place?
You want to generate energy in "unexpected places"? Why? I think you want to generate energy in places that need energy, no?
And why would you want to constrain the problem to use kites? What about all other energy sources? Is kites the best one? The most economical? Or is it just some engineering porn stuff?
Other problems seem to suffer from the same nonsensical approach:
"How can beams of light support the rapidly growing global demand for data?" Uh, what?
"How can balloons deliver the Internet to rural and unconnected places?" Who cares? How can paper airplanes deliver the internet to urban places? That's just a thought experiment with no motivation.
Etc. etc.
The space-X inter-satellite communication is photonics: beams of light.
So thats two things you pontificated on, you don't actually understand.
Loon is not 'blah, who cares' -Loon is 'wow: this is working, lets get this deployed' joined to 'we can re-purpose your model in LEO sats'
Do you seriously think "rural and unconnected places" is a who cares response? Here's a hint: go read the iridium story. People (with money) care a LOT about these spaces.
The point isn't that they're not doing "real work" or that what they're building doesn't work at all.
The point is about the approach which consists in engineering a "hard problem" by adding unnecessary constraints/hypotheses to a real (although often ill-defined) problem, and then claiming it's a marvel of engineering when a half-functional, non-viable solution is completed.
It's the XY problem applied to startups. Instead of trying to solve the problem (energy, connectivity, etc.), they start from a given technology, and try to shoehorn it into a problem. That's bad design, and probably bad engineering in most cases.
Numerous proposals, including from Lockheed, for balloon satellites: https://en.wikipedia.org/wiki/High-altitude_balloon#Geostati...
These aren't just "because we can" frivolous projects, these approaches have serious advantages over others, such that many are racing to commercialize it.
The former end up in a stable equilibrium with slow (if any) progress, the latter fail often but are important for fast-paced progress.
In other words, whenever you (the reader generally, not davidivadavid in particular) write "probably", imagine a Wikipedia editor adding [citation needed] after it.
Or that 5 billion people live in rural and remote areas?
* How can stem cells fight climate change?
* How can we use waste heat from nuclear reactors to promote gender equality?
* How can echidnas cure cancer?
The challenge is that many revolutionary products sound equally nonsensical at the time.
* How can we generate power homes using giant propellers?
* How can monkey viruses cure blindness?
* How can adding impurities to a conductor let it perform mathematics?
We don't seem to have a way to tell which crazy sentences will pan out until we try and see.
I think you're missing the OP's point: the point is that the sentence should read:
"How can we cure blindness?"
If it happens to be via monkey viruses, so be it. But you shouldn't start from "How can monkey viruses cure blindness" — it's not monkey viruses that are the reason to do research, it's blindness that's the reason.
If you problem is to generate power with kites, and you have built power-generating kites, then you won. But how good a power source is it in the grander scheme of things? Is it going to move the needle when it comes to CO2 emissions? Again, I don't mind the high school science fair stuff, but I don't think this has a high chance of being the ticket based on fundamentals.
https://news.ycombinator.com/item?id=22361216
In case you pay for ft and not Medium, like myself.
Direct link:
https://www.ft.com/content/56582032-528e-11ea-8841-482eed003...
The best we can do is get behind solar and wind and pray for a fusion breakthrough.
People accept trade-offs all the time. We understand some jets crash but we still don't give up air travel. We understand cars crash but we don't give up automobiles. I just think nuclear meltdown risk is overblown. If it happens four times a century, is that acceptable? I think it is. All things considered, how many people have died from nuclear meltdowns versus coal extraction?
Either energy collapse or war or both might result otherwise.
If we manage to avoid either (seems unlikely so far), then we just need to solve climate change and we’re almost at type 1 civ.
/speculation
Reducing population doesn't make it all that much easier to meet energy demand. The efficiency of power use and the types of plants we choose to build are much more important.
Although Alphabet apparently wasn't a great match, is Shell any better - are they really going to be an effective advocate for developing wind power as an effective competitor against their primary fossil fuel business?
See: GM with Maven, Ford with Chariot, BMW with Parkmobile.
"But Alphabet decided the opportunity to greatly disrupt to energy sector had begun to dwindle. While Makani at first felt it could provide a competitive advantage by using its technology on land and in shallow water — as well as deep water — advancements in competing technologies meant that was likely no longer the case.
“That’s very sad for Makani,” Mr Teller said. “But it’s really great for the world.”
An EU report published in 2018 concurred with Alphabet’s assessment. “The technology still has a long way to go before it can reach commercialisation,” it read, noting that challenges remained in demonstrating the reliable, autonomous operation that would make the concept viable."
It looks good for their image and seems a good fit for local conditions. In particular rigs stand-alone, so less concern with tether coming down on someone else.
Moonshots are inspirational, but sadly do not seem sustainable.
On the other hand, I wonder how many successful startups ex-X employees will have launched.
Facebook?
Learning what you can't do is as important as to know what you can do. I'm happy that Alphabet is doing these long term research projects and bets. >5 years I think, apart from government, only oil & gas companies project that long term, but they know much better what they will 'find'. What Alphabet does it moves the academic needle cause we learn so much on the way.
Maybe they can do better in publishing the results, cause that is an alternative to just shareholder value: society value.