In a boom cycle, it's not a great use of resources to try to really tighten up the manger/report ratios because the size of the org because the slack lets you grow faster. The higher the growth, the more managers at a more inefficient manager/report ratio generally.

Certain orgs in LinkedIn were growing really fast. Now they're not and they know the headcount for the next year, time to ratchet down the number of managers to get a more steady, cost-efficient ratio.

I think you're really under-estimating the positive, human value that can come out of what I'm describing.

If you leave the world of software companies you'll find that a lot of humanity is wasting huge amounts of time on tasks that could be easily be automated. My most recent experience was the Electronic Vehicle research space - I was able to rather straightforwardly reduce testing cycles for certain key components from 1 year to 1 month through some straightforward software and collaboration with some scientists.

Most of what I accomplished could have been achieved by the scientists if they had used something like Retool[0], but Retool is too sophisticated a tool for them to ramp up on. If AI could make Retool accessible to someone with the technical sophistication of Material Scientist who can write a little Python, it might greatly speed up the rate at which we advance EV technology.

The point I'm making is that making it easier to make products that are accessible means that it's easier to distribute the positive effects of innovation to the rest of society faster. If anything, there's the potential to lower profits long term because today creating a product that is both valuable and accessible is an incredible moat.

[0] https://retool.com/

I can now have 1-2 developers stand up ML backed services at a level of quality that a few years ago would have required an ML + engineering team to build along with an ongoing tuning burden. Now that the AI is "good enough" out of the box time-to-value has dropped, which also allows for more exploration.

One area I'm seeing a lot of traction at my company and amongst other developers: onboarding flows for complex products. LLMs are really great at taking a small amount of input from or about a user, walking down a decision tree, and creating some initial dummy data relevant to them to more quickly demonstrate value. You might not ever know chatGPT is involved but it doing wonders for quite a few companies' conversion rates.

The best tactical advice tends to from people who were recently in a very similar position as you, had had approximately the same goal you have, and either achieved the goal or failed. You want a sample of both successes and failures. The less recent, the less useful the advice.

The best strategic advice tends to be from people who have accomplished things you want to accomplish, regardless of time distance. Their tactical advice is less useful (because it is often dated), but the strategic advice tends to be better. You again need a large sample size to filter out noise.

I think we'll have to settle for judgment calls.

I tried going through some of the research we have on developer productivity a few years back and it's almost all garbage or is only truly applicable to juniors (e.g. when you're new you really benefit from quick feedback time on static errors).

The entire space suffers from the fact that good experimental design is impossible to implement with anyone who's not a college student (good luck getting professionals to follow your rules for months) and the curse of dimensionality from the need to disentangle individual variations, type of software development, management style, and a thousand other things to try and draw out a signal.

Sadly, many things and life can't be effectively measured.

Now there's a significant portion of the labor market that expects WFH, companies need to produce a policy on WFH/RTO instead of treating it as a non-decision.

Data gives no clear insight into which is better, which makes this a judgment call, and everyone with >5 yoe has enough experience in both modes that they feel qualified to make that judgment. Many think requiring some in-office time is superior. You can try and dismiss those opinions as "vague issues due to lack of water cooler conversations" but that's not going to actually convince anyone with the power to effect these decisions - even if you're right! You need answers to concerns like "virtual communication is too low bandwidth to build alignment on strategic shifts that are necessary for the company to grow to profitability" (quote to me from a director at a company with >1,000 employees).

My argument is basically: if you could have addressed those concerns, it would've happened during the pandemic. Manifestly, those concerns were not addressed in a satisfactory way. Therefore the only real resolution now is to wait 5-10 years to see if RTO/WFH is a meaningful differentiator for companies.

I think the hard reality is that companies need to make a thesis on the level of flexibility in remote/in-office work and commit, then 5 years from now we'll get an idea of what works well.

It is high, but not exceptionally high[0]. It is then worth remembering that the US overcounts spending on primary and secondary education relative to peer countries because we lack most of the social welfare programs that exist in peer countries and effectively funnel welfare services through the education budget.

As an example, if you have a severely emotionally disturbed child the school district might spend ~$100,000 a year to house the student elsewhere[1]. In Norway, this problem would be budgeted under social services, rather than education.

Numbers are further inflated by increased cost of living relative to most peer countries[2]. This is effectively a dead-weight cost for which we receive no additional benefit. Since Labor is the biggest cost for schools it greatly impacts cost of education.

[0] https://www.statista.com/statistics/238733/expenditure-on-ed...

[1] https://lao.ca.gov/Publications/Report/4110#:~:text=Special%....

[2] https://worldpopulationreview.com/country-rankings/most-expe...

An argument I find compelling here is as follows: sometimes I will look for something, and the world will tell me what I want does not exist. I will either build it myself or do without. If the state of the world changes, and a solution emerges, I would like that fact pushed to me (concrete example: new, more efficient battery chemistries for hot climates).

Waiting for everyone to pull slows down growth of whoever is innovating this new thing - which in turns hinders their innovation - which in turn slows down innovation in the economy.

There's a tradeoff here with all the bad things you lay out about advertising, but I'm not convinced I should prefer the more rapid rate of innovation.

One thing to ask yourself - is the company chasing growth? A company like AirBnB or Twitter could absolutely run on a much smaller team if they agreed to mostly abandon growth. They could focus on reducing tech debt, make architectural decisions that reduced flexibility in the name of reliability and lower maintenance, and then eventually lay off a huge portion of their staff.

Instead they have a growth thesis - that growth requires investment, having to favor velocity over unit economics, etc. The higher the growth target, the more inefficiencies you take on.

The inherent assumption here is that a trusted body can be established for a reasonable price that will accurately report these numbers so that the tariff can be enforced. This is a bad assumption both because of practical difficulties in many countries (try this in the DRC) and because one must often assume that the trading government is willing to subvert these tariffs.

In practice, the only practical solution is usually a flat tariff paired with diplomacy stating that the condition to revisit the tariff is building trust that these standards will be enforced on both sides.

This is what naive calculations say, this is not what a DEXA scan says. His body is actively reducing calorie expendature well below what you would expect.

I have tried taking over his calorie counting for a week to help him, as someone who manages my own weight successfully I don't think I would overcount by a factor of 2. The chance he ate calories I was unaware of during that week is approximately zero (highly controlled environment, never left apartment)

My father is obese, and has been for many years. At the weight he usually hovers at, his BMR appears to be ~2,500 calories. He can lose weight from here no problem eating ~1600 calories a day, not easy, but not to hard.

The problem is, that after about 15 lbs of weightloss, he plateaus. Dexafit scans confirm there's been a massive reduction in BMR. By the time he's dropped 20 lbs his calorie counts generally need to be in the 1200s to lose weight, which is just torturous. The longest he ever last at these really low calorie counts was probably a year, I'm significantly smaller fitter than he is and I could not have lived a productive life on the diet he was eating.

I'm not sure how common this kind of response is, but his doctor's didn't seem to think it uncommon.

Your resume now signals "white collar upper-middle class" and retail will assume:

- You'll still be looking for work when you join

- You can't/won't work hard

All my friends with PhDs in non-marketable specialties have faced this problem for a decade.

Here's a classic problem as an illustration: The storage cost of your logs is really prohibitive. You would like to cut out some of your logs from storage but cannot lower retention below some threshold (say 2 weeks maybe). For this example, assume that tracing is also enabled and every log has a traceId

A good answer is to run a compaction job that inspects each trace. If it contains an error preserve it. Remove X% of all other traces.

Log levels make the ergonomics for this excellent and it can save millions of dollars a year at sufficient scale.

Keeping engineers in the room then adds a positive feedback loop: as they understand the business better they start to offer up new opportunities to the business that have surprisingly low cost.

(I will caveat that for sufficiently solved problems you can generally involve software engineers less)

A lot of the revolution is coming from material scientists doing the work of creating data sets that encode their knowledge, better computational techniques in the field (e.g. DFT[0]), and democratization of petabyte scale data pipelines.

[0] https://en.wikipedia.org/wiki/Density_functional_theory

[0] https://www.ycombinator.com/companies/osium-ai

1. A substantial fraction of successful exits are now acquisitions. If you have ever been a part of an acquisition you know that unless the acquired company is a unicorn it is unlikely that you will see anything unless you are a key member of personnel. I've seen this happen to friends at >100 companies.

2. Long timelines to public complicate things. Options, even on a 10 year timeline, may expire before you can reasonably exercise them. The lack of cash can hinder life plans (children, house) and incentivise you to bet big. Each round requires the org to once again execute. If I build play a key role in making a business worth 100 million, and we raise at a value of 2 billion and fail to get there, my stock might now be worth very little.

The E(V) at larger companies is awesome by comparison.