Hacker News Reader

Hi Dan!

0. (We've met briefly; I work at a clean energy startup that relies on Utility API; I posted this.) 1. It is too late to prevent civilization-impacting warming, but it is not too late to have an enormous effect on how much impact happens how quickly. 2. "Optimism isn’t principally an analysis of present reality. It’s an ethic. It is not based on denial or rosy thinking. It is a moral posture toward the world we find ourselves in." [1] 3. Ken Caldiera thinks that the tool-building you have done is the right kind of work. [2] 4. Ya, HN has noisy nay-sayers. A lot of them simply have outdated mental models and excessive self-confidence. Can't do much about the latter, but posts like yours can chip away at the latter. 5. If you need a dose of fact-based optimism, check out recent editions of Chris Goodall's amazing mailing list. [3]

Your work matters. Keep on keeping on! Best, Sam

[1] https://talkingpointsmemo.com/edblog/observations-on-the-day... [2] https://kencaldeira.wordpress.com/2018/07/21/on-choosing-pro... [3] https://us9.campaign-archive.com/home/?u=a336c39e55a6260d59a...

Howdy! Thanks for the links. I didn't mean to come off as so fatalistic. I'm still very optimistic about our future, whether or not other people in tech participate (it will just take longer, and we'll suffer more consequences).

My point was to mostly to impress that problems exist outside of web tech, that web tech experts could help solve. The smartest people I know are in the oil industry. They know how to literally go into Nigeria, bribe the right gangs, and get the oil out. They are way better at "getting shit done" than anyone I know in silicon valley. These are the people those of us in cleantech are up against, so we're severely out-gunned.

We truly do need all the help we can get, even if it is from tech elites. I'm willing to make that deal with Worry Free[1].

[1]: Reference to Sorry to Bother You: https://www.theatlantic.com/entertainment/archive/2018/07/so...

IMO the problem is not optimism but either ignorance or fatalism and insanity:

- Most people know (at least I hope so) that drastic change is urgent. Why Hope Is Dangerous When It Comes to Climate Change http://www.slate.com/articles/technology/future_tense/2017/0...

- Most people are not willing to change and to let the government invest in change. Most people are used to inertia and more of the same. Drastic change and massive public investments in science and useful technology is not what most people are used to. Daily news bombard the people but not what is really important. Harmful military budgets are increased. Trump even proposed that NATO members should spend 4% of their GDP for harmful military. Animal products are still allowed. Politicians and their voters argue against electric cars because they are expensive and enough electricity is not yet delivered by wind and solar. They do not realize that all related industries and practices must change ASAP. Politicians and their voters enforce monetary austerity because public ownership and public investments for a different world is the worst that could possibly happen in their imagined world.

Annual global fossil fuel subsidies were amounting to $5.3 trillion in 2015 (6.5% of global GDP).

https://www.sciencedirect.com/science/article/pii/S0305750X1...

Why the 2C Target is Far Too High

http://www.climate-change-emergency-medical-response.org/2-d...

A degree by degree explanation of what will happen when the earth warms

http://globalwarming.berrens.nl/globalwarming.htm

That’s somewhat just the three modes of communication on HN. I see far too much elitism and general negativity here.

Politics are at play how ever. My understanding of renewables is they require multi-year, billion dollar projects. The desert I live in, for instance, has a political faction devoted to tourism development. And to develop geothermal here would come up against NIMBYs.

Personally I can program, build super computers, read and communicate math (bs in math). But I don’t know how I could really be hired to help.

As for salaries, energy companies represent some of the largest companies in the world. Just look at what energy money has done for Saudi Arabia. Or all the oil rig drillers who work 4 years and retire a millionaire.

Most of the supercomputers in energy are currently at grid operators (in California, it's called CAISO[1]). Lots of interesting neural networks stuff being done to drive the energy futures wholesale markets and grid services events. Those sorts of companies usually go to Grid <whatever>[2][3][4] conferences, so that's a good place to look for companies/agencies hiring (look at the speakers, sponsors, and exibitors). ERCOT in Texas has been particularly innovative with it's wholesale market operations.

[1]: https://en.wikipedia.org/wiki/California_Independent_System_...

[2]: https://sepapower.org/event-complex/grid-evolution-summit-na...

[3]: https://www.greentechmedia.com/events/live/grid-edge-innovat...

[4]: http://www.distributech.com/index.html

I bookmarked your comment, thank you so much!

Hey, this was really useful. Maybe you can help me (and many other potentially interested software engineers) by sharing another key missing piece: what are online places to lurk and gather initial information in this sector?

It is a bit difficult for most people to just know that they like working in some sector, or even to get in contact with the sector's community. Even more so, how to expect prospective software entrepreneurs to even get an idea to work on without contact and info. Of course, they can get out of their house, start going to industry meetings etc. but this requires lots of effort, money, and even moving to another city. Are there "energy HNs" to just follow for some time and try to get to know the business before a more serious involvement?

Maybe it's the assumption that salaries aren't competitive? Maybe it's because most business models in energy tech aren't compatible with VC? Maybe it's because you have no idea what specific problems there are that need solving? Maybe the assumption is that energy is super slow and bureaucratic?

Since you're asking, and I can only speak for myself, I don't like taking that path because at some point you'd need to communicate with some sort of local/national authority and/or politicians and I'd rather eat ear wax than go down that road.

Heh, seems like a good adaptation for the old saying. Politics and regulation is like third rail of tech, no one wants to touch it. But that's where the power is.

What was the original saying?

Entitlements, social security specifically, is the third rail.

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

> So why the the lack of interest from "high growth" HN crowd?

1) Because VC's want returns measured in months, not years. So VC funding is right out.

2) None of this is high-growth. You are selling into a market with a fixed budget. To win, you have to displace something else. That's expensive and difficult and doesn't scale.

3) You are effectively selling to the government with all the downsides. Worse, you are selling to a set of things that are government constrained without the upside of being actual government.

4) Nobody inside cares. There is no short term 15% bonus to most of the people with their hands on the levers. The people in charge won't move until there is so much residential solar installed and the power grid is in such bad shape that individuals start disconnecting from the grid.

1) You're right. Every time VC tries to get involved in energy, they lose their shirt (with some rare exceptions). Luckily, there are many other more established funding options in energy (strategics, grants, private equity, experienced angels, etc.). The same is true with other traditional engineering sectors (defense, infrastructure, etc.). High growth software != VC.

2) I'm confused, isn't this negated by market size? Energy market sizes are measured in trillions, so displacing incumbents is still an extremely profitable venture. However, what's lacking is the potential of a monopoly. Energy is a commodity, so there's always heavy competition, which I guess is another strike against the VC model, but not against the high growth possibility.

3) While governments are often customers, they are rarely the only or even the majority customer. However, what is true is that the energy sector learned long ago that lobbying earns a 22,000% rate of return[1], so they use that tool. A lot. That' doesn't mean it's not a place for software. It just means you have budget that in.

4) I'm not sure what "inside" means? Do you mean politicians? Energy CEOs? Software engineers? Mind explaining this point more? I can tell you that every leader I've met in the energy industry cares a ton about the incoming disruption, even if they are incentivized to do anything about it (e.g. your 15% bonus).

P.S. This was fun. Are you in the bay area? I'd love to grab coffee to hear more from your perspective!

[1]: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1375082

2) A lottery ticket payout can also be very large, but it still doesn't make it a good bet. The problem is that you will have to go a long time with zero income and then suddenly you hit the jackpot (or go bankrupt). That's not conducive to running a business.

4) The leaders only "care" about this disruption in the sense that they may lose their jobs when people start protesting or disconnecting from the grid. They certainly don't have any fear of a competitor showing up and blowing them out of the market. So, there is no incentive to make a move that requires them to spend significant money.

The fact that the big commercial property owners in Southern California were doing massive arbitrage with their energy storage during the last heatwave tells me that CAISO is WAY behind the ball. There is no reason that CAISO shouldn't be doing arbitrage to itself. This was also matched by information from some of the local universities who have cogeneration but don't actually push energy back to the grid because the grid operators we too far behind the curve.

I see these same issues in other industries, as well. Cold chain, for instance. Lots of people at the top want better tracking, but everybody's business processes are ossified in concrete. And nobody is authorized to actually spend money. And extra information will simply make life worse for the people at the bottom actually carrying out all the work. You will spend years throwing yourself at the wall in that industry until something finally breaks. Or, at some point, someone like Amazon says "Thou shalt track everything and report it to us" and suddenly the entire cold chain industry has a heart attack. However, the market can remain irrational longer than you can remain solvent.

I actually don't get up to the Bay Area much anymore, but if I'm ever in that part of the world I'll give you a plink.

Hacker News is interesting when actual experts decide to take part in the conversation. (Otherwise, it's no better than other places.)

So maybe this boils down to why there aren't many energy experts here? Where do they hang out, anyway?

"So why the the lack of interest from "high growth" HN crowd?"

I can't speak for the "crowd", but personally I've lost more money in solar stocks than any other sector I can think of. And I didn't even get burned by gambling on specific stocks like SunEdison, just by a broad solar ETF.

Just because something is the wave of the future doesn't mean it's a good investment. If you invested in the solar ETF called TAN[1] back when it came out ten years ago, it's down over 90%. It just keeps shrinking and shrinking, although it has bounced around a little.

Why should producing a commodity efficiently be lucrative from a profitability standpoint? Competitive markets should reduce the profit of a safe activity anyone can do to almost zero. And investors who are willing to invest without regard to profits should reduce investment returns to zero or below.

[1]https://www.nyse.com/quote/ARCX:TAN

Any product that your consumer not use at leisure time, get it's profitability reduce to zero.

My guess is that a lot of people have become jaded over time because energy contains such a high degree of business and politics.

Another hurdle might be that it's difficult for pure software developers to enter the field because most problems require a combination of hardware, software, infrastructure and regulation in order to be solved.

> So why the the lack of interest from "high growth" HN crowd?

Because it's not a problem that can be solved exclusively with software and marketed as a SaaS or eyeballs ploy.

There are plenty of people on HN solving hard problems but they are not too overlapping with the 'high growth' crowd.

This absolutely is a software problem.

Many utilities are interested in enrolling their customers in programs that would allow software to control the energy demands their customers put on the grid.

For instance, if you are a high-energy consuming factory, you may have a fixed amount of energy that you need to use each day, but have some flexibility as to exactly what time of day it is used.

There are companies right now writing software that takes into consideration all the constraints around the electrical usage of the various customers of a utility, and then vary the customer’s electrical usage in real time as the energy demand/production fluctuates on the grid.

Think of it this way. You can either put a battery on the grid to supply energy when the grid needs it OR, you can ask customers to stop using as much energy at that moment. Either way, the supply meets demand.

For reference: https://en.wikipedia.org/wiki/Virtual_power_plant

Although the problem you described can be solved with software, is it really "high growth"?

Put another way, can this grid flexibility problem be solved purely with a (currently popular in tech, especially among startups, as the GP alluded) cloud-based SaaS offering?

Or will it require some significant non-software portions to the solutions, such as co-locating hardware and/or high-reliability low-latency networking?

I realized my (own) sibling comment sounded like a bunch of rhetorical questions, since they were asked from the standpoint of my skepticism/suspicion based on what I read.

However, they actually are questions, and I'd appreciate knowing the answers (or best approximations) from an insider.

The wikipedia doesn't provide much insight into the technical underpinnings to draw any conclusions.

Perhaps there's something of an answer to the OC's general question, just in this inscrutability: the details are so foreign/unique but so important that software/computer people would have no intuition on how or where their skills would fit.

You seem to have missed the word 'exclusively'. Obviously software is an (important) part of a solution but it will need a lot of hardware and hardware knowledge in order to be able to be solved well. Your typical high-growth hacker wantrepreneur wouldn't touch a project like that with a 10' pole.

I think maybe the problem is that energy is broadly a solved problem as far as US is concerned. You guys have enough energy that you need and the means to pay for it. The world might move to solar/wind but it doesnt really matter to the average joe in the US because things are pretty good already.

Countries like India, Solar is huge. We don't get enough energy even in our richest cities. Most people cant afford to pay as much as they do. Using coal to fire electricity plants is clogging up the air. Solar has the potential to change a lot here.

It's always nice to hear from someone else in cleantech on here :)

I work for a startup that's focused on making it easier for utilities to integrate DERs into the distribution grid. We make software to simulate the grid, gain real time visibility, and unlock new business models.

I think that software interest in these kinds of industries will be coupled to a maturation of the software profession. When the software industry recognizes the need for professional licensing (like a P.Eng for engineers) in certain disciplines then energy will be much more appealing.

Yes, I agree that software engineering is still in its infancy compared to other engineering fields. I always keep in mind the final passage of They Write the Right Stuff (1996):

"And that’s the point: the shuttle process is so extreme, the drive for perfection is so focused, that it reveals what’s required to achieve relentless execution. The most important things the shuttle group does — carefully planning the software in advance, writing no code until the design is complete, making no changes without supporting blueprints, keeping a completely accurate record of the code — are not expensive. The process isn’t even rocket science. Its standard practice in almost every engineering discipline except software engineering."

https://www.fastcompany.com/28121/they-write-right-stuff

> it's a scaling problem, which means software opportunity.

From the perspective of someone who's not a software engineer but rather an ops engineer (a software concierge [1]), I find the idea of a scaling problem to be far more attractive than a software (writing) problem, for obvious reasons.

> Maybe it's because most business models in energy tech aren't compatible with VC?

> Maybe the assumption is that energy is super slow and bureaucratic?

The above two combined would be enough. As sibling comments have pointed out, the nature of the industry means that the players involved will be very large and/or political.

It's not that the above three qualities are somehow magical. It's that they're generally necessary (but not sufficient) to create that small, startup (usually high-growth) company environment that appeals to a certain subset here.

Focussing on just one aspect of such an environment without understanding the totality of the appeal (and/or effectiveness) can end up turning into cargo-culting, as with (if you'll forgive the caricature) large, bureaucratic companies attempting to attract talent by refurnishing their common areas with brightly-colored beanbags chairs.

[1] https://news.ycombinator.com/item?id=17484840

The issue with such software is that it's mission-critical. In other terms, it needs to be zero-downtime, bullet-proof, audited and certified by external actors and to be supported for the next 30-50 years.

Historically, all those limitations make the "move quickly and break things" approach popular in the current wave of startups impossible and the code with this kind of requirements has historically been implemented by large corporation already working in the industrial domain.

If anything, this feels a domain where GE and large contractors will be chosen to write mission-critical code over startups.

the real thing is that engineering is much harder and more expensive than software engineering.

> I have a startup in cleantech software,

I'd be really interested to find out more about what you're doing. How did you get into that line of work?

We're building an API to interact with utilities. Most interactions with utilities (data requests, interconnection requests, etc.) are still manual processes, adding significant customer acquisition and asset management costs and slowing down deployment for distributed energy resources (DERs) and energy efficiency (EE) vendors. Automating those interactions removes a huge friction point, allowing DERs and EE to scale much more effectively.

I got into this because I was a project engineer for a commercial energy consultant, and I was spending half my time wrangling utility data instead of actually working on designing and deploying projects. So I started UtilityAPI.

Are the utilities giving you access to their db? How are you translating API requests over to the utility if they don't already have an API? Don't take my questions as criticisms as I think it's interesting.

It varies, but by and large we do not have arbitrary access to utility customer data. For most utilities, customers are empowered to give written authorization for a party to act on their behalf to access/manage their utility data, so we have an online form they can fill out to grant that consent within the US-DOE DataGuard Privacy Guidelines (think GDPR for energy data). For some utilities, they are beginning to provide OAuth-style access for customers (I'm actually on the board of this "Green Button Connect" standard, and we'll be coming out with our own OAuth-style solution utilities can buy next year).

Overall, it varies from utility to utility and regulatory authority to regulatory authority. There's an incredible amount of politics and policy involved, but mostly it revolves around clear consent (which isn't ever a problem since the utility customer is the one wanting the energy audit or quote). The complexity of the regulation/policy is what prevents a good fit for most VC models, and other financing options are more experienced and comfortable at this than VC.

Thanks for the response. Didn't immediately see it.

I agree with your sentiment, HN is notorious for armchair-quarterbacking with this subject.

Do you know of any cleantech software that take on remote workers?

Just adding my voice to "you're not alone!". I've been helping build a software solution to make the indeterminacy problem of solar a somewhat predicable one [0] by working to build an API that focuses on solar radiation and PV power production anywhere in the world updating every 10-15 minutes. Software will play a huge part in address these issues with the grid and clean electricity supply, and due to the recent economics of solar and wind, this is only going to accelerate. So I have also been surprised by the lack of competition in this space given the massive industry, real need of solutions, and the scale of money involved [1].

> Maybe it's the assumption that salaries aren't competitive? Maybe it's because most business models in energy tech aren't compatible with VC? Maybe it's because you have no idea what specific problems there are that need solving? Maybe the assumption is that energy is super slow and bureaucratic?

I was convinced that this is where I wanted to take my career from reading Bret Victor's piece on a technologists role in address climate change [2], originally thinking that this will never be a huge industry but more and more after reading this piece, realizing how large the opportunity is for software and technologists. The domain is so large there are so many angles, however, I think why VCs might be largely passing on based on your last point "the assumption is that energy is super slow and bureaucratic" as it is somewhat true but there seems to be a growing appetite to 'get things done' but I think this is more recent(?).

Something else that might be seen as a risk (certainly occurred to me) is that targeting an industry with so much money, one of them might just solve this problem themselves and they are already an insider to the industry. As much money as VCs have, it's pretty hard to outspend energy giants trying who really want problem X solved.

> Because it really is too late.

This scares me as well, from what looks like a start of a biodiversity crash [3], islands starting to disappear [4] and projecting out what all this means in over the next 50-100 years, it is "too late", however, I'm always reminded how things that look bad can always get worse so it's worth trying to do what we can now.

It's funny (in a sick way) how the joke that VCs always use the tag line of "to change the world", yet these problems are staring us in the face and that same drive seems to be largely missing, maybe cause "playing a small part to make things less horrible in the future" doesn't have the same ring to it..

[0]: https://solcast.com.au/

[1]: https://electrek.co/2018/01/23/tesla-giant-battery-australia...

[2]: http://worrydream.com/ClimateChange/

[3]: https://www.theguardian.com/environment/2017/oct/18/warning-...

[4]: https://www.forbes.com/sites/trevornace/2017/09/09/new-study...

> the scale of money involved [1]

And yet in that very example, that scale of money is involved for a battery solution. Notably lacking is how software, or computer technology of any kind, contributed to that dollar amount.

I don't doubt that opportunities do exist, and some have been outlined elsewhere in the thread. I'm just unconvinced that the overall money involved has anythign to do with the money available to pure software solutions.

Even in computing, not all problems can be solved with software.

Right, maybe that is a reason it has been avoided by VCs? Not many 'purely' software options?

Most of these problems probably do require a hardware component for the solution. Battery arbitrage (quite a bit of software involved here than you might think, forecasting generation, demand/supply etc) is probably on the extreme end of a software:hardware ratio (towards hardware investment). Software I'm helping to build is probably on the other extreme and we've still got some (very small) hardware components that we are directly involved with but also a very large one, latest generation geostationary weather satellites, however these are externalized (eg we didn't put them up there, maintain them etc).

There is plenty of scope of problems in the energy world, but yes I'd agree that money for 'pure' software solutions is probably a very small part of the pie. I think this comes down to fact that the electricity grid or its components are at the center of all of it, so integration points have to exist. Maybe also seen as high risk, energy supply is ignored when working well but becomes very obvious when it's not.

Glad to see in this thread there are more people on HN than I thought working in this area, a lot of effort is still needed.

> that is a reason it has been avoided by VCs? Not many 'purely' software options?

This does strike me as reason enough, yes, not that it's necessarily rational. Although there are certainly VC dollars available for non-software startups, including consumer hardware (which strikes me as exceedingly risky) and biotech, when it comes to software, there seems to be a trend to keeping it pure, even avoiding buying computer hardware and staying on cloud, despite the tremendous cost premium (especially when scaling!).

I suspect the overall instinct is from the enterprise (or even consumer) software days, where the engineering cost to produce the software is completely decoupled from the revenue it can generate. Add a significant hardware component and the VC imagines they'll be permanently coupled, with profit forever limited by that hardware cost. Of course, this is also true for SaaS, perhaps more so when the hardware is rented at a markup from someone else. I remain mystified, but it's mostly speculation, anyway.

didn't know 'cleantech' was a term, but it is: https://en.wikipedia.org/wiki/Clean_technology

What sort of tech work specifically? What languages are used?

The answers to your questions are basically "all of it" and "whatever makes the system work." The Department of Energy is pouring millions into 'smart grid' R&D, which includes everything from good old cybersecurity all the way up through nuclear power research, stopping at coal, diesel, steam, solar, and wind along the way.

> pouring millions into 'smart grid' R&D

Unfortunately, that essentially contradict's the GP's "isn't a research problem anymore" assertion.

Similarly, millions, even tens of millions, isn't an impressive amount of money by VC-funded high-growth tech startup standards, so that also belies the notion that it's not just research any more.

I feel like you're looking for a hole to poke. It seems pretty obvious that R&D funding is still good even when existing technologies are good-enough-to-deploy-in-mass.

Ford got past the R&D stage for the Model-T a long time ago, but they continue to pour money into R&D to make better cars. The same happens for energy technology. We have clean energy systems that can compete with fossil fuels today, but that doesn't mean we shouldn't keep seeking new innovations that make them even better.

I wasn't suggesting that research can no longer exist, just that the OC's whole point was that this field is no longer about just research and attempting to open up the discussion to a broader interest group, of software engineers, based on that premise of mass scaling.

Since the GP was a direct reply to the comment with that premise, asking about the kinds of work and technology, it seemed to me like a borderline-disingenuous to respond with something related to research and not mass scaling, especially since that response was "all of it".

Here's a comment on what specific software needs there are in distributed solar[1], and here's a comment on where to look around for interesting energy companies that may be hiring software engineers[2].

[1]: https://news.ycombinator.com/item?id=13250336

[2]: https://news.ycombinator.com/item?id=15127154