It's the opposite. An LLM is better at CEO stuff than working code. A good developer + LLM instead of CEO can succeed. A good CEO + LLM instead of developer cannot succeed. (For a tech company)
Is that a fact? I mean, see the linked article; even the company whose whole business model lies in convincing people that that _is_ a fact is kinda saying “yeah, perhaps not”, with vague promises of jam tomorrow.
My father was and none of his bosses showed any signs of disrespect during his 40 year career with the company.
My bosses treat people like shit and they get bonuses for that. I know for a fact that none of my peers will retire from the company they currently work at. They will find some sort of excuse to fire them way before that age.
Of course not; the people who are the most pro-union know nothing about them. If he experienced how dehumanizing and power hungry unions are he wouldn’t support them.
It’s not about whether an argument can be made where becoming a for profit aligns with goals. The more general issue is whether one should be able to raise money with the pretext of a non-profit, pay no taxes, and later decide to take it all private.
In that case, why shouldn’t original funders be retroactively converted to investors and be given a large part of the ownership?
I don’t understand why all these comments are about who buys or doesn’t buy at Amazon. The article is about unionization and strikes. I expect a conversation about the merits of unions and their negotiation tactics. In my opinion, events like these will just accelerate job elimination. The goal of a logistics company is to be reliable. Humans are unreliable and more so when they are purposely and collectively unreliable. I’ve lived in a country with very powerful unions and it sucks— miss every 5th flight because the union decides to strike.
> I don’t understand why all these comments are about who buys or doesn’t buy at Amazon.
The logical connection is as follows: Bezos decided to optimize everything to its limits, including human behavior, to the very limits of law. To literally track every movement of employees and abusing the power the company holds over them. This is an ongoing process that we are all painfully aware of. Because of that, there is a growing negative feeling towards them that causes people not to give them their money. That's why instead of unions we are talking about boycotting Amazon.
> In my opinion, events like these will just accelerate job elimination
Sure, probably. The jobs that can be automated will eventually be automated. But while they're still needed, I'd hope they have some basic protections and decent wages.
If my online shopping costs go up 0.5% but now a thousand workers don't have to have the mental stress of "I really need a pee break, can my metrics take the few minute hit this week or will I lose my job and go homeless?", I'll take that trade in a heartbeat.
Most companies’ largest expense is labor/wages (usually somewhere on the order of 25-50%), and profit margins are usually on the order of 0-5%. Increasing pay or benefits substantially would increase costs by a lot more than 0.5%.
Sure, 0.5% is hyperbole on my part. Sorry. It's not like a 20% bump in costs for these workers would result in a 20% bump in prices or anything like that.
Their margin on that retail item is probably 30-40% of the cost of that product though. Let's assume the workers' benefits and wages in question here are 35% of Amazon's costs. If there was a 20% increase of that labor cost, that's going from 35% to 42% of the total share of costs, or an increase of 7% of the total costs. But that's 7% of the 30-40% of their markup. For a product with a 40% markup and they were to just pass that entire cost along to the consumer, it's a 1.6% increase in price.
So like in this hypothetical, which is not anywhere near real numbers for Amazon, we could give these workers a 20% benefit bump for increasing prices 1.6%.
I don't understand how you got down from 7% down to 1.6%. I think you are not counting COGS as an expense, and that's where the error comes from. If your assumptions are otherwise correct, I think you'd get roughly a 7% increase in prices to sustain the 20% bump in wage expense.
I am including COGS in this; that's the retail price minus the markup. The markup is all the rest of Amazon's costs, of which I agree wages for workers are probably somewhere around 30%ish.
Are you suggesting the warehouse worker's benefits and pay is really 25-50%+ of the final purchase price of the good?? That'd be an extraordinarily high amount of cost.
Yes, though I am including delivery and support staff with warehouse workers (basically everyone who is working 'on the floor'). More than half of their employees are categorized as "laborers and helpers", and there are a number of other categories that seem similar. https://assets.aboutamazon.com/64/79/d3746ef14fd99cc6be94532... (I only found this after your latest comment).
The largest categories of employees tend to dominate most companies' cost structures. I would like to run some numbers to see what the likely distribution is here, but the annual filings are quite sparse (in terms of income statement details), and I don't have the time to do an extensive analysis.
The merits of unions, not always but their goals are, multiple. a/ compensation and working conditions leverage in negotiations b/ structure that can directly address all sort of issues the employer don't care about and isn't obliged to deal with.
That it makes your service experience particularly painful is exactly the goal, bad (or better, no) customers experience hurts the employer, guaranteed.
If a certain arrangement makes it more likely to write bad queries, and it requires extra care to write optimal queries, then it’s a worse interface to a database. I bet for really database intensive applications graphQL adds more work than it saves.
You're talking about the implementation of the protocol, right?
That is a good implementation of it, called GraphQL Yoga[0]
However I'm concerned there is a slight disconnect here. I'm saying that the technical specification of GraphQL does not lend itself to being bad, rather its the failure of developers to really understand its purpose and what its for (its a giant aggregator, with various ways to optimally aggregate things together, depending on what is optimal for a given problem set)
For that, I recommend becoming more familiar with the specification itself[1] because thats what I'm talking about. The specification (and thus its technical nature) doesn't prescribe anything regarding how you get data on to the graph. Many people equate GraphQL with database problems[2]
This doesn't mean I don't understand that GraphQL has shortcomings, but all approaches to APIs have short comings. I have found GraphQL has the least amount
[2]: Common complaint I see all the time. I find it stems from a failure to understand how the entirety of GraphQL is meant to work, and some of the mechanics within. Like when to appropriately leverage DataLoader[3], for instance.
> You should maximize your income while minimizing efforts that take you away from living your life
You should maximize long term net worth, which often is in conflict with maximizing near term income. One can argue net worth is also correlated with compounding knowledge which is itself anti-correlated with "minimizing effort".
This whole viewpoint of work as a simple transaction hurts employees' long term prospects.
As someone that has been on the hiring side, I always prefer to build a full time team, even if that means paying for someone in my team to learn something new.
The part that consultants don’t talk about is that you have to pay them to learn your code / company too. Nobody can just jump in and add value immediately. So you’re paying for onboarding hours, you’re paying for other employees to educate the consultant, and so on.
Likewise as an engineer, I think the most interesting projects and companies err towards primarily full time people. So if my goal is to work on the most interesting projects, I’d want to work full time.
A lot of consultants, for example someone helping with SOC2 compliance, is mostly copy pasting a large document with recommendations and moving some paragraphs around based on interviews with the team.
Not different from your average lawyer helping with an estate plan or will. The reality is that for many things you do want the cheapest lawyer, because the project is simple and repetitive. It’s the uneducated that pay more. So yes for a murder trial I want to pay a lot for a lawyer, but for small things I don’t, and you shouldn’t either. I simply don’t buy the analogy.
Finally, while consulting might be good for cash flow to hours-worked efficiency, it’s not clear it will make a person wealthier than other forms of income. I’ve met rich employees, founders, but not so much independent consultants.
I've been on both sides of the table for both full-time and consultants. The first rule of humans is that they vary widely within categories (even if there are statistically significant differences in population average), so I urge people to take all of the anecdotes on this page with many grains of salt.
That said, here are some of my own, condensed across experiences:
- An executive wanted to use ML, couldn't get enough political capital to build a full-time team (and didn't have the knowledge to interview candidates anyway). A very small consulting firm, sourced from high personal trust networks, got an MVP going. That led to a FT team getting built. Code was a mess but started the flywheel and after several years everyone agrees it was a great ROI
- A large company had a tiny team using a technology that would be standard for FAANG but they had no support or training. Zero progress for about a year. Not important enough to throw huge money at but that link in the chain was becoming a bottleneck for the core product. A consultant came in explicitly as a multi-week trainer, and but the team just never got up to speed. Consultant's code examples grew until they were just direct bugfixes, and though nobody was happy, at least the crisis was resolved.
- A product finally reaches product market fit. Executives decide to hire only after the team is drowning in tech debt. But it has to be full time, and it has to be junior. It takes months, then requires training. Tech debt is twice as bad as the original crisis level when the new hire becomes value positive. For some bizarre reason he leaves after exactly 2 years.
Many consultants have a branding problem: Their biggest victories are fixing a disaster that their clients would rather not talk about publicly. Hard to put on a billboard, if you ever want work from that client again!
I nearly always got a job offer at the end of my contracts. I never accepted them because the terms were always poor, they couldn't match my rates. Not only that but being a contractor offers some distance from the internal politics.
Usually the internal staff knew the business well but were weak on the technology they were using. No amount of training seemed to fix that.
The only downside for me was the travel which was why I used the downtime between jobs to pivot to developing my own product.
ah man, this so much. I've had the honor to work with very good freelancers and couldn't imagine they would've stayed as long as they did if I wouldn't have firewalled them off. I hope someone will do the same to me when I venture into the freelance world.
B/c internal politics is where it constantly breaks for me. I can only play along for so long.
Regarding the last point: maybe you're based in the US and you know some rich employees there, but I don't know a single one where I live. Consulting is very hard - for the reasons you mention, and others - but in some places it has a much higher ceiling than being an employee.
For those who don't know - aerhardt lives in Spain.
The US seems pretty distinct in that you can be a laborer and have quite high incomes. It's not unheard of to see people in major cities within the US having 7-figure incomes and still being an individual contributor in their field. Most of the "consultants" I've heard of outside of the US are only "rich" in that their incomes are high for their region (charging HCOL US rates while living in a place that is LCOL) or they're not doing labor anymore because they own a consultancy firm. In the US, I think consulting isn't worth it unless you are extremely lucky. You're more likely to see high income by trying to join FAANG.
I've been one of these people with the 7-fig incomes in the US. Unsurprisingly, you'll see a lot of data points like mine here because HN is primarily focused on the US audience since YC is a US (primarily silicon valley) company.
> Nobody can just jump in and add value immediately.
That depends on the project and how much domain knowledge the consultant/contractor has. I've certainly been able to contribute on day 1 on several of my gigs.
> I’ve met rich employees, founders, but not so much independent consultants.
May depend on the market, and what you mean by 'rich'. Consultants rarely get stock, but they often make out far better than employees. I've met very few employees that got anywhere near 'rich' but consultants in London (for example) can certainly become quite wealthy.
> As someone that has been on the hiring side, I always prefer to build a full time team, even if that means paying for someone in my team to learn something new.
> The part that consultants don’t talk about is that you have to pay them to learn your code / company too. Nobody can just jump in and add value immediately. So you’re paying for onboarding hours, you’re paying for other employees to educate the consultant, and so on.
as somebody who's been on the team, i agree on this. in my previous job my skip-level management insisted on handing off some crucial infrastructure work to external consultants.
results:
1. we on the (internal) devops/sre team did not know how parts of our infrastructure worked
2. when changes were needed, the first lenghty step was to figure out what was built by the consultant, and how (the manager did not like to pay the extra hours, of course)
3. when help was needed to troubleshoot something along with developers, we (devops/sre) ended up playing the role of messenger, and stuff that could have take anything from 2 hours to two days ended up taking up to two weeks
It was a frustrating job, so of course some people (me included) ended up leaving the company.
I still see the value of consultants as people that can bring the knowledge you need when you need it. However if they're going to work on something long lived, the manager must be enforcing some proper knowledge transfer during the consultant employment and at the end of it.
> The part that consultants don’t talk about is that you have to pay them to learn your code / company too.
Why should they? That's a given. Of course they don't know your code. But it's likely they've seen more code than a typical employee and therefore have more experience to contribute. A consultant is a serior employee sans the commitment.
To your point about time, consultants don't need 1 v 1s, quarterly / annual reviews, etc.
I understand why your cynical. I've seen some half-ass'ed consultants as well. But I've seen just as many stale, complacent, and complicit employees.
> So you’re paying for onboarding hours, you’re paying for other employees to educate the consultant, and so on.
You also pay for the psychological friction and pressure that the consultant has to suffer from the fact that customer expects them to be geniuses with magical powers and the project managers planning the time line according to that.
Yeah, it's tough to know who can actually deliver.
I know one guy (a consultant now) who I would trust to Navy SEALs parachute into almost any technical problem and come up with a real solution. Any part of the stack, from building a physical hardware test bench, debugging firmware compiler bugs, desktop software development, to a scalable web stack. These people exist, and they're not cheap.
However, it's worth noting that no one with skills like that works at Deloitte.
Deloitte does not hire great people. They would have to pay them great and corporate managers would rather cut themselves an arm than paying well a good employee, it is a matter of principles.
Exactly. The guy I'm thinking of has a small number of very happy clients and he can choose who he wants to work with. It's funny that at the high end, it seems like things get easier. You don't have to haggle over an hour or two. You name your price and deliver super high value to people who are happy to pay.
It makes me think of "The Wolf" from Pulp Fiction.
"It's about thirty minutes away. I'll be there in ten."
If you're lucky and the consultant is experienced and has narrow scope, they can provide a lot of structure. They will also push back where applicable. If they have a wide scope and/or they don't have much experience, you are 100% correct that they have to figure it out like anyone else.
> Nobody can just jump in and add value immediately.
Hmmm, I do this all the time. I've taken a bunch of bad data analytics, ML code from my clients and quickly given them insights that more than paid for my work.
A true expert can often come in and "fix" things quickly.
The problem is that in today’s climate, FTE is not really what it means. A FTE can also vanish at anytime, except it actually will ask you to pay more. And when a core tech leader left, there are also very few people can actually do any nontrivial code change to the repo.
Sure, sure. I think the comment was meant as FTE as strong evidence of continuity. As an employer, you shouldn't rely on any one FTE employee staying indefinitely or giving you very long notice before leaving. You need to think about which employees have key knowledge, experience or skills and what it would mean for your group if they were to leave. As an employee you can't rely on the group existing indefinitely or keeping you on indefinitely. Keep your network, skills and resume in shape accordingly.
Consultants can get that if they work for a firm. I get a 401k (+match), employee healthcare plan, and bonuses, for example. The firm doesn't have any stock for the other though.
You know the truism about how the best leads are often reluctant ones?
Feels like there's a parallel in consulting. The best ones will coach you to replace them with full-time hires. That being said, there are a lot of occasions where reality means that contractor may be the better option and a win.
Yeah it’s probably true. I explored this during a therapy session but nothing particularly insightful came up.
I grew up in a family of entrepreneurs (extremely small scale brick-and-mortar solo businesses in my home country, barely making ends meet) and I clearly remember my father and uncles in my childhood being very proud about “not having to work for someone else”, probably something subconsciously stuck.
To be clear, those same people are the ones who are now telling me to absolutely not quit my corporate march due to my demonstrated earning power, and to toughen up and suppress my feelings of unhappiness for as long as this money train can continue. My mother is a bit more romantic and she’s telling me to set a deadline in my 40s, which is still way, way too far.
Regarding monetary donations, if you believe institutional misuse outweighs efficiency gain, then your area food banks are in a much worse shape than my local area, which I hazard to guess is merely average.
Food banks can do so much more with cash, such as buy fresh produce from local farmers, than they can with your unwanted groceries, usually canned goods. Or if you're buying food specifically intending to donate it, do you really think the average person reading this (and 50% of you reading this are below average) is a better shopper with their $100 than the food bank is with $10,000+? Instead of shopping, spend that time with them, and still donate the same amount you would have spent.
I was surprised when I spent my own time and found out how much further they can stretch each dollar, and often provide healthier food.
