Ask HN: I automated part of someone else's job in my free time – how to proceed?
I've been working at this company for some time in an administrative role. As part of my job, we – and many others – work with data which gets aggregated and refined by a centralized team of specialists. This process is only partially automated and therefore still quite heavy-handed. As I understand it, significant time is spent on the manual part of it all.
Out of curiosity I started building this small project in my free time just to see if I could automate it myself, since most of the data required is available open source. There are still many rough edges but what I have now is already superior to what we have at work.
Now I'm wondering about where to go from here - so far I haven't told anyone. Strictly speaking, this is not part of my job and I am "poaching" in someone else's area of expertise. I like my job and don't want to make enemies. On the other hand, I would love to see my solution put to use and maybe even enhance and/or expand it jointly with IT. However, this is code I wrote in my free time and I am rather unwilling to give it away without any benefit.
Any ideas?
130 comments
[ 3.3 ms ] story [ 154 ms ] threadThe cynical me says they weren't smart enough to automate away their own work (so they could be assigned more, different work), so eff 'em, right? That is, until someone else automates away your job.
The business me says, you found and fixed an inefficiency, so capitalize on it. While you're at it why not add in extra shadow surveillance "features", and you've got yourself a unicorn. Get that VC money!
The "boss" me says, uhh... you did this work on "your own time" (as if there is such a thing) but with proprietary knowledge of my systems, "open source data" notwithstanding. So, yeahhh that work belongs to me, and I'll monetize it for myself thank you very much. That work product is not yours. Get the eff back to work already.
The "Overemployed" me says, hey, with this automation, maybe I can "work" two jobs simultaneously and pull double salary for no net increase in work. Get that easy money! At least until my employers discover the ruse and they both fire me!
Sometimes you just need to take a calculated risk and see where it goes.
Sort of besides the point but I wish someone _would_ automate my job so I can find even more interesting things to work on; I’m trying and I haven’t figured out how to yet. Yes, this is obviously only a position someone can feel if they are blessed with either lots of ability or economic security, but could we change that?
I know the problem is the resulting economic hit and not the having of a job or not, and I wish we had more of a safety net and support for people shuffling to better economic activities rather than being stuck in a local maxima with potentially disastrous switching costs. Things like UBI, public option healthcare, free state sponsored college degrees, and government sponsored vocational training have a decent record, and I wonder how we can scale them up past things like the GI bill.
What makes you think you would work on "more interesting things" as opposed to "nothing". That is, why would your employer keep you on as opposed to firing you.
But, to me, that would’ve been preferable to doing busy make-work, or worse lobbying to keep the inefficiencies that provide me the busywork I use to justify a salary. Once something becomes automate-able, I think it should be. I have never seen software people at the big co FAANG companies I’ve worked at get let go because their core job was automated, only reassigned out of dead end roles that never would have gotten them promoted. I understand this position is only possible to hold alongside all those privileges.
> But, to me, that would’ve been preferable to doing busy make-work
Good news. You can totally stop doing busy make-work and find another job, start consulting or start a business right now! You don't have to wait to be fired.
Yes, if you work at a large software shop, they may reassign you. Or maybe not. Apparently even FAANG is trimming workforce.
As an aside about large software shops reassigning, I’ve found at large software shops a big part of the role is making up meaningful work for yourself that helps the firm, and that’s a large part of what is actually rewarded (as opposed to fixing bugs or maintenance). I think a lot of people fall into the career trap of needing their immediate manager to narrowly define what is useful exactly. Usually when you are made redundant nobody reassigns you directly they just kinda ask you to figure it out and give you a couple weeks to a month to talk to teams with headcount and pick one with a problem area you think you can make progress on.
In theory, looking at a complete society's productivity, unemployment is a great thing, because it means little work is needed to sustain a specific level of wealth.
The problem is only one of distribution.
I believe strongly that the current economic system is effective primarily at concentrating wealth at the top and using the general population as a pool of cheap desperate labor. I believe we can actually generate more wealth if we abandon intellectual property and develop technologies specifically to free people from toil in an open, community-oriented way. Of course this will be a fight because those who hold the reins of power today have no interest in diluting their power.
"But automation is expensive" - well engineering development is expensive. Which is why I think we should spread the costs out as much as possible - make everything open source, so none of the work is repeated, and anyone in the world hungry to get in on the action can contribute, instead of secretive siloed orgs spending a hundred million on some secret sauce for one single application, without sharing the benefit of that work for all the adjacent applications. Instead smaller investments distributed throughout the world will lead to large scale innovation of the crowd, the way we saw 3D printer development explode AFTER the patents expired. Because 1000 people working on an idea get way more done that one well funded organization in many cases. This is exactly the kind of innovation that patents extinguish, and worldwide agreements ensure patents have widespread enforcement.
We all need the same basic necessities, so there is so much overlap, that these systems built by the crowd could rapidly mature.
Eventually we could envision a system that can function with just 10% workforce participation, as people basically volunteer for maintenance and development. Before then we need to educate as many people as possible on engineering, robotics, and computer science.
But the end goal of a world where a highly automated "means of production" is held in community ownership is I think very important. Working in a world of privately held companies with closed source engineering leads to VERY different engineering requirements, and fails to realize benefits of community based engineering development. There are people all over the world that need what you are developing, and some of them have the time and energy to make it their own and contribute their own spin. But only if we make that community.
I have tried to discuss these concepts on my youtube channel in this latest video, though it was unscripted and I'd like to produce a clearer scripted version in the future:
https://www.youtube.com/watch?v=Sx_2-xwWJUk
EDIT: For additional motivation, I highly recommend this talk by Jenny Odell called "How To Do Nothing". I feel so strongly that we are wasting ourselves away at jobs that could be automated away, and this talk really captures some of what the alternatives could be for the human psyche:
https://www.youtube.com/watch?v=izjlP9qtmBU
I don't think you understand the concept of risk, particularly as it pertains to things like war and armed conflict.
Defense spending always looks excessive and unnecessary until you need it. Saying a country overestimates how much defense spending they need is like saying you don't need car insurance because you've never been in an accident.
Wars are similar to car accidents. We're really bad at predicting when we're going to get into one and it can be a whole lot worse than we ever could have imagined.
Additionally for most western countries, the U.S. military acts as a fail-safe of sorts. If Belgium, for instance, underestimates their defense needs they know at some point the U.S. will jump in to help them if need be, which is largely what's happening in Ukraine right now.
By contrast, if the U.S. underestimates its defense needs and suddenly finds itself under attack, say by the Chinese, who exactly would come and save them?
But all this talk of military spending is orthogonal to my original comment. I am saying we would be better off if washing machines were open source, and the workers at the washing machine factory would be better off if the organization was cooperatively owned and managed. I never suggested abolishing the state. The comment mentioning the military was in error if they thought I was talking about making my own country. You can do what I propose inside the USA, and the government can keep taxing exchanges to fund the largest military the world has ever seen. I have opinions on that, but those opinions don't relate to the idea that non-military technology should be open source. Obviously militaries prefer secrets.
I will say, it is interesting that you say I don't understand risk. The USA does not have the largest army humanity has ever seen in order to manage risk. The USA has this army to maintain power over other countries. Belgium as you mention relies on the USA for some extra defense protection. But this also means that Belgium cannot oppose us. We have covered the globe in US military bases so that we can maintain control. This strategy goes far beyond managing risk. If we merely wanted to manage risk, we could have avoided treating every country south of our border like shit, fostered healthy relationships with them, and after decades of healthy relationships formed a military alliance with them. Then that organization would defend the Americas against foreign invaders, not the USA.
Last year, I would have responded to your position with this quote from Ruth Bader Ginsburg's dissent in Shelby County v. Holder, "t]hrowing out preclearance when it has worked and is continuing to work to stop discriminatory changes is like throwing away your umbrella in a rainstorm because you are not getting wet." and reminded you why you're dry. But this year, with the umbrella being on full display and not only stopping rain, but also stopping a genocide by the murderous LGBT-hating Siloviki class ruling Russia, I'm just kind of confused about how you could argue that we've got way too much umbrella.
Sadly, this automates away the #1 source of meaning in one's life. "I made it this far, in spite of all the adversity" is what most decent adults lean into to motivate themselves out of their beds in the morning.
I really don’t think that compulsory labor is the best way for people to find meaning. We’re not going to sit around on the couch forever feeling so miserable that we can’t go out and do anything. We’d be so free to explore what we want! Sure, people raised in a society that worships labor would feel unsure of their footing for a while, but I don’t think that’s a deep part of the human condition. People absolutely would still work. Imagine creating a theater production. All the work on planning, set design and construction, etc. None of that may pay a living but it’s still meaningful work that most people can’t spend time doing today. In a world without compulsory labor more meaningful work will be found.
In the US at least, we’re going to get a point in the next few generations where it’s going to get harder and harder for many to find meaning in life. Increased automation, increased climate change, decreased financial security for people who expected to be secure.
It’s going to be rough.
"Decent" as in: we believe that character, including work ethic, can make or break a life.
And people can still work for more if they want. I just don’t think that makes sense as a focus for everyone, because it devolves in to what we have today, which is a mess for most people.
But ultimately what is decent is up to everyone. I just think open community innovation is a better way to manage progress than what we have today, and would benefit everyone more than what we have now.
We aren’t exactly meeting everyone’s needs to the maximum we could today.
EDIT: I just think open community innovation is more likely to produce what more people want and need. I think it will actually produce more innovation than the current system. Remember that the sole function of a patent is to prevent innovation on top of existing ideas by third parties.
For another example, see this real world practical example in India where a single facility gives out 50,000-100,000 free meals every day. There is a real economics to this, and it’s something I want to study. I presume they have a multitude of funding sources including donations of money and product, probably some government support for the free meals they send to schools, and volunteer labor. But they have an organization which across India gives out over 1.2 million free meals a day. It’s well worth studying how they make that work, but they certainly do make it work.
https://youtu.be/qdoJroKUwu0
But imagine you are part of a co-op that owns an automated farm. How do y’all determine who gets what? Well that’s up to the members to decide. But I think it’s an eminently solvable problem. On a larger scale different co-ops might trade using money, or might have sharing agreements between co-ops. Again, it’s up to the members.
And then for things that are reusable, libraries are a great source of inspiration. The Oakland public library has a tool lending library where you can borrow all kinds of common household tools for free. I think there’s a great number of possibilities for how to share wealth without the exchange of money, and in other cases money can be handled at the organizational level.
And what you mightn't see is the threat this poses to the evidently unfair, but also evidently well entrenched status quo.
I think it’s important to really understand what is at stake. The current structure of society is such that everyone is under the thumb of assholes today. Better to have an asshole as another member of a collective than an asshole as your boss.
I do certainly understand the threat this poses to the well entrenched status quo. My strategy is to start small so the big players don’t believe you’re worth stopping, and then once you’ve proven your method works you’ve already picked up enough momentum to keep moving despite opposition. But in the end this is a project that will take lifetimes, and may never be achieved. I’d still rather work on this than design some robot designed to make the lives of very comfortable people slightly more comfortable, while simultaneously ensuring the CEO gets a new vacation home in Colorado.
Look around, your eminence. Most of our species prefer not working at all, and if we must, prefer expending the least mental effort.
The problem is that we want work that is accessible, and interesting. It should be right near the edge but firmly within our abilities to be satisfying and meaningful, and we have problems when it deviates from that significantly.
It’s possible that for wide swaths of people that edge won’t keep up with the edge of what is not automate-able, but I think a) accepting that someone people will have to be supported out of work and b) others can have their personal ability edge pushed further with vocational training is both good for those individuals and good for society.
In the words of Chris Farley, "sounds like a personal problem."
As for yourself, you have an important decision to make. Are you a wolf or a sheep?
Since when do wolves kill all their pck members so they can hunt alone?
If they are receptive ... I am happy to look around for any possible solutions that might help you. Will let you know if I find anything that might interest you.
If my boss / other team sound comfortable I might share snippets or demo a possible solution a week later.
Otherwise if the response is not encouraging I will keep it to myself.
After you tell your manager, you won't be in control of what happens next. Consider whether or not you should just keep this to yourself and treat this as a fun learning experience. Perhaps you can gauge their potential reaction by asking, hypothetically, whether or not they think working on automation might be a good idea.
I've done something like this before, but with the knowledge of the other people on the team, because we didn't like the manual work we were doing and felt that it was taking time away from other things we needed to do. Hopefully that is the case for your colleagues, or this will not be well received.
If you frame it as, "hey, I did this thing that I think might solve a problem and here's a demo," it might work out, or it might not. You could get a promotion out of it, or maybe not. If it's a good idea and the company doesn't want to do it, or doesn't recognize you for it, then you don't want to work there in the long term anyway.
Some of the comments are basically advising not to rock the boat. Anyone in 2022 who thinks their current computer-related job will not become automated or obsolete at some point in the near or distant future is living in a dream world. Nobody is entitled to a job doing tedious work.
(Quick edit: trying hard to leverage this to get any special credit or $$ or other goodness out of it is a non-starter. You did this with company data probably on company time with company tooling. It's for them - and that's ok! You are demonstrating that you are there for the good of the team on this one. Many points awarded.)
Telling your manager feels like the right first move and will likely help make it the best it can be, but the biggest thing is going to be the culture of your team/company.
(p.s. If it doesn't work out and you like working with data...I am hiring https://jobs.lever.co/bateau).
As far as ownership of this project goes, you need to think carefully before making any sort of claim. Depending on the language of your employment agreement, the IP of this project likely belongs to the company because it's based on their systems. "I did it on my own time" probably isn't going to hold up in a lawsuit either. And companies tend to have a lot more money to spend on lawyers than individual contributors, especially in cases where significant aspects of their operational model is threatened.
No no no. Go to your manager first. Otherwise that other person might tell your manager, who will be very displeased about being blindsided.
If you know your manager would react badly, even if they heard about this directly from you, then you already know what to do: keep this to yourself.
Nobody wants to lose their job especially these days.
Yes OP will make an enemy even at the suggestion this code exists. Plus OP will be labeled by employees in the company as someone to not trust.
They might start asking you to help with other things. Others will hear about it, not as you're trying to step on them, but that you're making their life easier. You could become the "go to person". Then you might get some time to study on the company dime.
> As I understand it, significant time is spent on the manual part of it all.
Get more details about the problem space before doing anything else.
What I've learned with automating my own tasks (I am in an administrative role myself), is that there's usually not much of an advantage in telling it. Which is why nowadays I don't tell anyone and just use the added free time however I want. (well, part of the reason is that I don't like my team enough to share it with them, as harsh as it sounds. I liked sharing with my prior team because there I knew that everyone was doing their part...) The incentives usually are not set up well for automating your own (or your colleagues') tasks. Would probably be different if you were in a manager role, though.
Someone I used to work with had what I would consider a tedious portion of her job automated away by an officially sanctioned project involving multiple people in multiple groups. She thought everyone was trying to get rid of her.
I can't tell you how to communicate your intentions so they are not taken the wrong way, just that there is a reasonable likelihood that they will be.
Sibling comments address other pitfalls.
My advice: if the coding part is what you enjoy, pursue that. Go become a software engineer.
"Hey Boss, could I get a $50k bonus and a promotion with 20% raise if I automated 30 hours of work per week that someone else does?"
If Yes - get it in writing and say "Here it is!"
If Not - do nothing.
IME, 4 out of 5 times it’s complete wrong, feelings will get hurt. Engage your manager if you feel comfortable, but I would scale back the ambition/expectations and start from there.
Acknowledging that, you have a bunch of really bad options and few good options.
The absolute safest thing you can do is to keep it to yourself, and to use it when you can to make your own job easier. If you’re able to find time to use the tool to load your own data because the team that owns this process is overloaded, then they’ll appreciate the relief, so long as your tool doesn’t break or violate their standing policies.
The best advice in this thread is going to your manager and asking for time to develop and demo an automation solution. No one wants to hear that you’ve already built it because that raises thorny political and management questions. The best you can hope for is a sanctioned meeting where you can demo the capability to your manager and the manager of the other team.
If you’re fortunate enough to get authorization to proceed, make sure to tell everyone that your solution only handles the common simple cases and that the tricky special cases still need manual review- an 80% solution. The achieves two goals: allowing you to get away with having bugs, and allowing the other team to believe that they are more specialized and, hence, more valuable now with your solution than before.
Most software engineers, for obvious reasons, fall into the category of the former. They will always be a minority of the general working population.
The toiling majority generally does not take kindly to having been made redundant. Proceed with caution.
Even if your idea is good, logical and beneficial someone in management will take it as a slight that you had a good idea they didn't. Maybe you will have receptive management who are decent human beings, but in my experience you will be lynched for daring to step outside your lane.
Proceed with extreme caution. Your good feeling vibes about the effort you put in will NEVER be received the way you want. Management will see this as a slight and a threat.
Imagine the person who's job you've automated decides to quit.
"Hey boss, instead of taking the time and money to replace so-and-so, how about you just increase my pay by (percentage) of what the salary would be? I have a solution to this staffing issue you might be interested in."
I don't think this would happen to you, but you might lose some other folks their job. Something for you to consider.