Calibrate Before You Accelerate: Bias Toward Action in a New Role
tucker.wales
tucker.wales
Another time we got bought, merged into a new company, their CTO was our new CTO. In one of the initial meetings said "things won't change much" and, naturally, we didn't believe. She moved slow on all the things, methodical. Spent some time with each component team. It was months before we started making changes to better mesh with the new company. Very little chaos. Still admire that management style.
The right way to make those changes, or at least one way and I don't know a better one, is to perturb the system and see how it responds. This is necessarily risky and uncomfortable, and choosing the right things to perturb and the magnitude of those peturbations is a very, very tricky skill.
But ultimately if you're dealing with an opaque system, which most companies are, and you want to make it a more transparent system that can change, that risk and discomfort are necessary side effects.
This is of course very situational, and the methodical method is by far preferred, but if you'll permit a metaphor: you get dropped into an unfamiliar cockpit, and the controls are labeled in some foreign language. The plane is crashing, or at least you don't know if it is. What do you do? You wiggle things and hope that, in doing so, you don't crash, and you take very careful note of what happens.
While most companies have elements that are hard to see clearly [1], I would hope 95% of what the company is doing would be transparent to the CEO.
The CEO should easily be able to tell which divisions/projects are profitable, which are on track to hit their goals, how those goals combine into a coherent strategy, which areas are getting the most complaints and refunds, and so on.
[1] intangible concepts like 'innovation' and 'culture' for example
It might happen that this is the correct approach, and how to slice things up is often an eternal argument both on an organizational level and technical level.
Unfortunately that gives a lot of power to the middlemen, whoch they might use against you, but that's how the cookie crumbles.
We accept best-effort from CTOs, but that leads to an industry that is a bit like a comet with a few success stories and a huge burning tail of leadership where they have limited to no ability to understand whether they are on track to achieve their goals and those goals don't come together into a coherent strategy. There is a strong argument that many of the success stories are only good relative to companies that failed even worse than they did and we all had to pick one of these bug-ridden software packages.
It is a lot like the era of the manufacturing industry before the statisticians moved in around WWII. We haven't managed a similar moment in software yet.
That's why founder types have such a massive advantage. They've got a much better mental picture of what's beyond the door.
Sometimes the plane is going to crash and the best you can do is a soft landing.
I'm a fan of Chesterton's Fence [1]. Put effort into understanding the organization and why it's architected the way it is before making changes.
The other part of that is ensuring that it doesn't fall disproportionately on certain sections of the company, and that you provide compensatory benefits like, say, comp overtime, extra vacation, bonuses, whatever is possible.
In the matter of reforming things, as distinct from deforming them, there is one plain and simple principle; a principle which will probably be called a paradox. There exists in such a case a certain institution or law; let us say, for the sake of simplicity, a fence or gate erected across a road. The more modern type of reformer goes gaily up to it and says, "I don't see the use of this; let us clear it away." To which the more intelligent type of reformer will do well to answer: "If you don't see the use of it, I certainly won't let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it."
— https://en.wikipedia.org/wiki/G._K._Chesterton#Chesterton's_...
Except in many cases it is your responsibility to ask. If something has indefined legal status, and no documentation, it is a major red flag. Very often corruption, drug or people trafficking or other shady stuff.
If you want gate in middle of the public road, show the paperwork! Or I will call authorities!
Maybe there’s a reason you’re on fire. Until you can say for sure, I will not put it out.
I will just be here roasting these marshmallows on your burning flesh. :)
Something which, as soon as you discover it, you should immediately respond by trying to remove it, without asking questions about where it came from or what its purpose might be.
I also double checked on gptzero me and it 100% agrees with me.
I'm curious to know though whether or not the "author" used Gemini. I've been using Gemini a lot in the past year, and the writing sounds exactly like Gemini. But it's possible that all the models sound the same.
Edit: I'd like to add that I still liked the article and agree with it, and in general I find Gemini's writing style to be quite enjoyable.
In fact, I think they've got so much more management in their training set than a senior engineer that I expect they're going to be a better source of material than the author.
The blueprint with the “LOAD-BEARING WALL!!!” includes nonsensical labels like “Dining Broom” and “Witchen”, and I am supremely disappointed that there isn’t a bubbling cauldron or pointed hat to be found.
Yeah, nobody took one second to look this over before posting it directly from the slop machine =(
I visit HN because it's supposed to surface interesting, novel, engaging, challenging articles. LLM output is the extreme opposite of that.
AI is a tool. It outputs things only when asked to. Asking it requires skill. Saying “This is AI generated” is like pointing at furniture and saying that “power tools were used to build that” or at a program and saying “that’s not assembly, a modern garbage-collected language was used to build that” or at a meal and saying “this was cooked in a kitchen with a modern temperature-controlled oven” or at a paper and saying “this was obviously typeset in LaTeX, not roff or plain TeX.”
So what? Give me a shop full of power tools and I can’t build any furniture, give me Python and a bunch of libraries and I can’t write Mercurial, and give me any LLM and I couldn’t have written this piece.
Does the conceit come from the idea of “some LLM wrote this, so anyone could”? This is like walking through a modern art museum and seeing some obvious-looking piece and saying “I could have done that.” Well, you didn’t.
This is not entirely accurate if we consider agentic loops.
> Asking it requires skill.
Considering the realm of writing a blog post instead of programming, what is the skill? I suspect there is extremely little. But I could be wrong, tell me what skill you think went into this.
> and give me any LLM and I couldn’t have written this piece.
Prompt your LLM of choice to generate 100 articles about different aspects of joining a new role, then you won’t need to provide the insights into the specific situations the author described. All you need to do is pay for the tokens.
> Does the conceit come from the idea of “some LLM wrote this, so anyone could”?
Yes.
> Well, you didn’t.
Do you know that? What’s stopping someone from asking an LLM to generate an entire HN front page worth of articles, and reading that list instead of HN?
This really is a problem with the health of the site in my opinion. If there are a flood of AI generated articles in the top spot, the value proposition of HN as a site is diminished.
I'd argue that counts as using AI as a tool for conveying my original thoughts.
I believe there is a societal problem here, not specific to your particular actions. The expectation that communications should be “cleaned up” and “professional” in the first place comes from a bygone era. In the 2010s and earlier removing all typos, fixing wording, formatting the text and adding images were all signals that someone put extra care into sending their message. But in the present that just indicates use of AI.
You mentioned in another comment that your hand-drawn images would be “far worse”. Leaving aside whether any images were needed at all, a hand-drawn scribble in the 2010s might be unprofessional but in the present day there’s an AI generated image on every visible surface. A hand-drawn image, even a bad one, would actually make you stand out.
Even if you want to present the cleaned up version, you could also provide the original audio.
I do think it's interesting that some people are immediately turned off of something because of the whiff of AI even if they like the actual idea or content.
Will likely experiment with different ways of writing in the future!
Even if the ideas are of your own merit, you can't expect readers to always do the literary surgery it will take to identify that.That's extra investigative work and assessment that was never a part of anyones reportoire before all this took off. But at this point i don't think you should be surprised that AI writing isn't met with open arms, AI writing has become a pain point for a lot of people in their personal lives and their work lives, a lot of organizations are wounding themselves as people throw their ideas around curated in AI and everyone's eyes are glazing over. It's become synonymous with "this guy does not understand the claude proposal he just sent me". And the choice is yours if you still want to use it knowing that, I just don't think you should be surprised if it doesn't get the reception you hoped for.
Plus, you should lean into whatever your monologue is to write. Even if you're afraid it's not pro enough, like. That's writing. It's a journey it has crossover with your ability to communicate it defines you as a person. Keep sharpening that tool and you have untold ways to convey ideas exactly the way you want people to experience them for the first time.
Also, you engaged with my comment, not I with yours, and I attempted to converse with you, but you then exited superciliously (below) without engaging with or even acknowledging my point. You don't need to say goodbye online if you don't want to discuss what I'm saying.
The first paragraph is probably not AI generated
"A bias toward action is a superpower only when applied correctly - here's how to frame it as moving decisively after building context, not rushing in before you have it."
...followed by many, many bloated words that can basically be summed up in two words of old Internet lore: LURK MOAR
I agree with the article too, but I didn't have to read that many words to know. "Don't just start doing things because you want to show off, and understand what value you're providing" is how I'd phrase it.
> I distinguish four types. There are clever, hardworking, stupid, and lazy officers. Usually two characteristics are combined. Some are clever and hardworking; their place is the General Staff. The next ones are stupid and lazy; they make up 90 percent of every army and are suited to routine duties. Anyone who is both clever and lazy is qualified for the highest leadership duties, because he possesses the mental clarity and strength of nerve necessary for difficult decisions. One must beware of anyone who is both stupid and hardworking; he must not be entrusted with any responsibility because he will always only cause damage. [0]
So reframed, TFA argues that one must be "hardworking and clever" instead of "hardworking and stupid".
[0]: https://en.wikipedia.org/wiki/Kurt_von_Hammerstein-Equord#Cl...
Reputations are hard to earn and easy to destroy. Trust building at any new institution takes a lot of time and effort and can seem annoying. But just doing solid work and communicating about it consistently will make sure that like minded people notice you, vouch for you and then give you more opportunities.
https://i.pinimg.com/736x/a5/b3/10/a5b31033a487595f913639c35...
- what is going well,
- what is not going well
-what do hope that I will fix
-what should I do
-what should I not do
I put those question in the agenda for the meeting invite so they could come prepared. Then we took the conversation from there.
I put it all into a google doc, then synthesized it into summary that hid the identities of the people who said it and used it as my guide for the first few months.
Props to the author if they simply don’t care though.
CMM says that an engineering team whose processes aren't written down is level 0, and writing them down, even if they are batshit, gets you to level 1. Which leads to a fun bit of catharsis with the older employees where they get to say things like, "and then a miracle occurs".
Also writing things down gives someone else a peek into what's going on in your head and they can correct bad assumptions you have before they get cemented and take more effort to dig out of your thought processes.
So I always start with fixing the documentation, the runbooks, the CI process. It's a deliverable you can engage in without breaking production, it demonstrates mastery, it fixes a pain point that the more professionally mature members of the team care about, which gets you brownie points with the right sort of people. And it makes it easier to onboard the next person, or pick back up a project that has been on the back burner for several quarters.
TODOs emerge from the documentation or get explained away as unnecessary or wontfix. By the time you're touching something you have a better idea of why things are like they are, so you make up for 'lost' time.
This is action, just not action that can tarnish both of us by shipping bugs to production.
Edit to add: I used to take ex coworkers out for coffee or beers around their last day and ask for a rundown of all the reasons they left. What you generally find if you let them keep talking is that they will run through the problems in reverse chronological order. The last thing they will mention is almost always what a joke the onboarding process was.
Last straws are the most recent thing that set the person off. All the straws before it add up, and if a person is already questioning the maturity of the organization on day 2 on the team, then I believe that multiplies your turnover rate. The longer you can go before a new employee says "what the actual fuck", is a multiplier on how long they will stick around.
Effectively, I think "the first 100 days" rule of thumb goes both ways. You get 100 days to show you're useful as an employee, but everyone already on the team is being held to the same yardstick by new employees. And if you're hiring at a high enough rate, having 20% of the team think the old employees are a waste of oxygen is not good for consensus building.
It takes a really good teacher to be able to explain as easily to an intern as to a principal, and often the intermediate people have to take the beginners aside and reframe your explanations anyway. Whether you witness it happen or they do it where you cannot see, it happens.
Making the ultimate visibility tool isn't just helpful to me as a newcomer. It may be useful to many veterans and people who haven't had the time to make this tool for themselves.
But I'm also somewhat biased towards laziness/observation.
As a result, after 15 years as a software engineer, I’m genuinely considering leaving the industry altogether because the only roles available to me are ones where I’m expected to deliver features rather than organisational change and growth. It’s like my heaps of experience have navigated my career into a cup-de-sac, and the only way out is backwards. I’m so jaded. Hopefully it’s a phase. I need a coach. Help.
It’s critical to start showing some Phase 3 impact, even if not with a sledgehammer, before peers quickly loose faith
a new hand who doesn't have to wonder what the body is doing will be much more effective than one feeling around for other limbs.
In such environments the people that do demonstrate a bias towards action tend to fall into one of two camps: those that wish they hadn't and those that are chasing attention.
The corporate developers who do have an overwhelming bias towards action, not the sociopath attention chasers, just end up contributing to open source projects unrelated to employment tasks.
I must admit I’ve always considered this to be self-evident but maybe I’m the odd one out.
I’d be a terrible blogger.
But yes, the article can basically be distilled into just Chesterton's Fence.