My boss does this for "estimating" software project schedules. He built a goddamned spreadsheet* where he will rate the entire project on a scale of 1 to 5, with 1 being trivial/quick-win territory, and 5 being extremely labor-intensive.

Two problems with this approach as used at my job:

1. He assigns the ratings before requirements gathering has even started (if they ever get documented in the first place).
2. He bases the final deadline around the calculator spreadsheet, and sends that date on to the business partners/project stakeholders within the company, and they usually pass it along to upper management.

So, by the time we finally get requirements together and find out, oh, shit, this is actually way more complicated than a 2.71828 or whatever, the stakeholders have already told the Senior VPs of Thought Leadering that my team will be done by a specific date. The week before that date rolls around, boss goes into a panic, demands that I work on absolutely nothing else as I'm being pinged daily to put out random bullshit fires on other projects that were rushed through implementation before I even worked here. Between that and the low pay, I start really strongly considering pulling a no-show. I stay up late a couple of nights, project gets finished. Rinse. Repeat.

I envy the dead.

*: No, it's not a Monte Carlo simulation or anything that fancy -- he just multiplies the complexity rating by a set number of labor hours, and doesn't bake in additional time for risk mitigation. They promoted his ass because this is so scientific and data-driven. Edit: and no, there isn't a more detailed breakdown/implementation milestone schedule somewhere further down in the estimate. It's literally "I feel like this is a... 2. You have a week. GIT 'ER DUN!"

Yes, I agree. I think it's only 3x, 4x at best according to my gut feeling

Not to mention he has failed to even outline why that's the case.

Big Yud is literally the founder of an assigning-random-values-based-on-gut-feelings religion. His essay on Bayesian reasoning has given perhaps thousands of blog-reading nerds irreversible brain damage.

These are the people that decide your wage

Just slap some rss over that bad boy, how bad could it be? (boy I sure hope there aren't other, unstated requirements :no-thoughts:

Also hire me to be one of those 10 (nothing funny will happen as I am a deeply serious person)

He has to think that evil AGI is the most significant threat to humanity or he doesn't get his stipend.

as someone currently doing contracting to clean up a big database which has been mismanaged and poorly maintained this entire twitter thread gives me a professional panic attack

im breathing into a paper bag rn

I found a YouTube link in your comment. Here are links to the same video on alternative frontends that protect your privacy:

• yewtu.be
• inv.nadeko.net
• yt.artemislena.eu
• piped.video

CW - SA: He also wrote a short story where people in the future/aliens are appalled that we (i.e. present-day humans) considered r*pe to be a bad thing. You did not read that wrong.

He does not do anything other than write fanfiction

i forgot he did stuff other than write fanfiction

He doesn't, really. It's just that some of what he writes is openly called "fan fiction" and some of it is in disguise.

This is a real guy, sadly. Very well known and influential in certain Effective Altruism/Pseudointellectual Silicon Valley cicles

Hard to say, but there's a decent chance he isn't afaik he doesn't really know anything.

I think the issue comes when you look at all the functionality that is much more nuanced than just the bare technicals.

So he's right that you could make Twitter if you just don't implement 99% of the features that make Twitter, Twitter. Not to mention all the workers that work on the non-product side... All the various infra teams, security, abuse, etc. etc.

bruh come on...

Yeah, it basically comes down to a complete lack of comprehension for how big something like twitter really is. On the surface level the functionality is pretty simple. But there's so much else going on that nobody sees, and a whole heap of it will be interconnected.

Twitter web, twitter app for ios and android, twitter api, advertising, content monitoring, content storage, caching, serving, twitter for businesses, content algorithms, accounts, privacy features, user settings, theming, ui, ux, embedded content. That's just off the top of my head. I'm sure a lot of these huge companies could be a bit leaner than they are, but usually the size is somewhat warranted.

This guys whole thing is just making stupid takes based on absolute surface level knowledge of things and sounding confident enough that people buy into it.

With 10 engineers one should be able to set up a Mastodon instance and scale it

A Mastodon instance is used by, at best, a few hundred to low thousands of people, and is going to be small and relatively obscure

Twitter is used by millions, is the preferred quick communication tool of tens of thousands of companies, and is one of the single biggest presences on the net. It'll take far more than 10 engineers to keep it running when it gets randomly DDOSed for a laugh by some bored teenagers, where a Mastodon instance either wouldn't even be a target or would just accept going down temporarily

mastodon itself has like 900 contributors tho, with 23 fairly active contributors. the distributed nature of it means that rather than just having 10 engineers, they need at least 1 maintainer for every instance. there are currently ~10,000 instances. so somewhere around 10,000 or more people are keeping it running

Mastodon demonstrably does not scale to twitter numbers. Even without feature parity it would be unusable.

Granted, twitter's unusable now, but it used to be usable.