This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
Could AI be the next big thing without impacting the economy and labour market?
Computers revolutionized the construction industry. CAD software makes it far easier to draw buildings and share the drawings. Phones makes communications vastly easier. Instead of a worker getting stuck or having to physically find someone they can make a video call. Manuals and documents are freely available online. Online shopping makes order parts cheaper and easier while allowing builders to press prices. Accounting, scheduling, recruiting sales and other supporting activities are easier with computers. Even on the construction site computers control machines. A modern truck is full of software.
Yet the productivity in the construction industry has flat lined and is if anything declining. Land prices can take some blame but renovating a building has not become cheaper.
Could we see similar effects with AI? A company in 2035 has completely automated customer service, AI drafts contracts, does sales and codes. We may have self driving cars and humanoid robots. Yet we might see barely 2% GDP growth and no real boom in productivity. Why has the tech sector revolutionized work without dramatical increases in productivity and can the results be better in the coming 20 years?
Some of programs are surprisingly quite limited. Chat GPT cannot download videos off youtube for example. Or remove text from images. Some random website can do it, but not 'state of the art' AI. Instead it shows a guide on how to do it with python . So basically it's like google, instead of actually automating said task. It also runs into meager data usage limits when performing computations , like trying to to solve 4x4 systems of equations (it will run into limitations when trying to solve more 4 of these matrices in 5-10 minute interval), and makes mistakes with other operations such as complex logarithms. Again, crappy websites can do this without limitations. It excels at rewriting and text though. For a free program it's not bad, but does not live up to the hype either. So I think it may not be the economic gamechanger as some expect.
More options
Context Copy link
People moved from productive roles to non-productive roles in response. HR wrecking your ability to hire. Endless meetings where nothing happens. Work that should and could be done in weeks takes months because the people on the other side are just lazy and everyone is too polite and unbothered to insist on a reasonable schedule (why be rude and damage relationships when there's all this money floating around).
Construction is an especially bad case, I consider it to have been deliberately sabotaged by vested interests, people whose entire job is to prevent development and construction with inane zoning or regulations. It really isn't that hard. Singapore has seen construction productivity rising. China can build large apartments in weeks, there are videos of it happening. Potholes that would linger for aeons in America disappear quickly in Japan.
Productivity in terms of 'wealth created per person actually working' has risen rapidly.
AI can raise productivity hugely, providing that implementation isn't sabotaged by the usual suspects. But it will reduce the number of producers and create vast opposing lobbies of angry & unemployed + wreckers and saboteurs. Thus I suspect we will see both productivity stagnation and productivity explosion, just like in the construction industry. Software companies may become massively more productive, only to hire many more charismatic, respected, dignified, useless management staff and thus keep their productivity where it was. Or they might just become massively more productive and skip the bloat. Countries can choose whether to do things efficiently and cheaply or whether they'll pay more and wait longer for inferior products. Of course, making the wrong choice will eventually lead to having sovereignty and wealth stripped away.
Coming from a history of working for small businesses to my first "real" salaried job at a large company I'm presently going crazy dealing with this. I'm current in "training" three weeks into the job and I've done next to nothing (My signature accomplishment so far has been performing a BIOS update on a coworker's laptop which just happened to fix his docking problem, sparing him a potentially weeks-long computer outage.). IT is either snowed over with work, incompetent, or just not a priority because their onboarding program doesn't exist and response time on tickets is glacial and requires escalating up the chain to get even basic shit done.
I was issued a company laptop without being told the username and password to log into it (had to call IT for that). The instructions for setting up the company phone didn't totally work and I had to figure it out myself (I'm a talented or at least "willing to Google it and try" or "capable of installing and using an easy Linux distro" user at best. I'm a car guy, not a computer guy.). It took me three weeks, multiple tickets, and multiple conversations with bosses to get the login to the dispatch software I'm supposed to use. I still don't have a login to the company intranet (Allegedly HR never created the account on their end.) and my email account is getting the wrong terminal's mail.
Call me crazy, but this is stuff that should've been done on day one, or week one at most. I should've been issued my devices with a piece of paper that had the relevant login credentials included and accounts already set up. My previous employer was a bunch of clowns that had their company phones shut off for a week due to non-payment and even they were capable of this. I'm assured that this is perfectly normal for the company, it's frustrating but par for the course, I'm not expected to actually contribute for a few months, and so on but I can't shake the fear that I'm totally wasting time in which I should be learning how to do my job and thus am going to wind up being thrown into the fire with little other than knowing the right people to call for help (because I've really spent most of my time hobnobbing and doing my best to present myself as eager to learn and do in lieu of actually doing anything) while I learn things on my own the hard way, under fire with expectations once the honeymoon period wears off.
More options
Context Copy link
More options
Context Copy link
Is this true? Granted, it's hard to measure productivity. But US GDP per capita, adjusted for inflation, has increased by about 40% since the tech boom began in the mid-1990s.
I wonder what are some good metrics of productivity.
Overall GDP can grow thanks to a couple of sectors while most others are left stagnating. Without looking too much into the numbers, one can guess finance and tech had this role in western economies in the last decades.
More options
Context Copy link
More options
Context Copy link
I saw an Nvidia presentation where Ai can be used to simulate an assembly line or an entire factory or warehouse, including even modeling the physics of the entire process of moving and assembling goods. Surely, there will some productivity gains from all of this.
Why? Going from paper to digital was a much bigger step, and it created almost no productivity gains.
Didn't it?
In college I had a part-time job with the facilities engineers. They'd digitized the blueprints for every building on campus, plus the full history of change orders. Before that, they had to go down into the halon-equipped archive and pull out file drawers with the originals. Surely that led to some productivity boost.
A couple years back, I was talking to an elderly woman who had worked in Saudi Arabia in (I believe) the 80s. She did payroll for an American-run hospital system, and oversaw their transition from bags full of paper money to checks. It sounded like a real quality of life improvement for the employer. Employees were a little more reluctant, but today, paychecks are ubiquituous. Except they've also been superseded by faster, self-documenting digital finance.
Then there's programmers. Even mirroring your hard drives has got to be more convenient and more scalable than a couple extra filing cabinets of punch cards. I don't even want to think about how the programmers of yore attempted version control. The productivity gains from digitization were obvious.
I suspect these generalize to most data-based industries. We're just more likely to take the improvements for granted.
That's my entire point, It did... but it didn't. The gains in productivity from paper to digital were massive, they should have allowed us to carve out entire swathes of administrative bloat out of our systems, but instead the opposite happened. I have my own anecdotes from older people living through the digital revolution, and Nybbler got my point instantly - we just came up with more meaningless paperwork to fill, to compensate for the gains.
More options
Context Copy link
More options
Context Copy link
Right. Making paperwork easier to handle just meant the major constraint on the increase of paperwork was lost, so paperwork increased vastly.
A key factor is how productivity improvements are handled by legacy companies vs new entrants. The digital revolution probably didn’t change much for Ohio Widgets PLC, with its large unionised workforce and complex compliance requirements. For Shenzhen Widgets LLC, on the other hand, digitisation is essential for its ability to take customised CNC machined orders from anywhere in the world, translate them into Mandarin, and have them shipped anywhere in the world in 5 days.
If that's so much better, why haven't we seen a massive increase in GDP or quality of life?
Have you seen pictures of any large Chinese city in 1990 and compared it to the same city in 2020?
Nope!
More options
Context Copy link
More options
Context Copy link
We have.
And far more for the Chinese.
Ok yeah everyone is talking about China, that could be fair! But China was also industrializing as well right? In the last century they were half peasant farmers.
@doglatine @SkoomaDentist @greyenlightenment
More options
Context Copy link
More options
Context Copy link
China has seen a massive increase in GDP and quality of life, as has the small segment of American society that has leveraged integration of new technology either directly into their industries or been able to exploit its integration elsewhere. And everyday American consumers have also seen a big increase in purchasing power thanks to cheap manufactured goods from China; the cost of textiles, consumer electronics, and household appliances has absolutely plummeted in real terms. However, lots of American industries have remained extremely stagnant in terms of productivity because they didn't/couldn't/wouldn't integrate new methods, so their purchasing power relative to the rest of the country and the rest of the world has declined.
More options
Context Copy link
Have you looked at how much China’s GDP has risen?
More options
Context Copy link
what do you mean? by almost every metric life has gotten better the past century? GDP keep chugging along too
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Construction productivity hasn't flatlined. Construction workers have cheap and easy access to power tools that significantly outperform the plug in tools of 2000. Anyways your other points don't really make sense.
Construction was likely largely switched to CAD by 2000
Except for trainees, who make up a minority of construction workers, they should be able to do their jobs without calling for help.
Not necessarily considered as inputs when computing labor productivity for construction.
Your unsourced chart is also worth less than the pixels it's printed on, unless you can provide measurement methodology. Looking up the numbers published by BLS https://www.bls.gov/productivity/highlights/construction-labor-productivity.htm shows an example of garbage in garbage out. It shows a huge spike in productivity before 2008 which your chart doesn't. Unless your methodology is perfect, the number likely reflects macroeconomic trends more than the actual efficiency of each worker.
That got me thinking. Even if a AI robot couldn't be out there swinging a hammer with me a camera that can take in the situation and tell me what to do, pre order the parts,, then walk me through through the work would be really hand in construction, and working on vehicles. Probably cut my 5 trips to Home Depot to replace the water heater to 2. Or that time I wired in heater take into the wrong run of Romex and now it only turns on when the downstairs kitchen light is turned off...
More options
Context Copy link
More options
Context Copy link
I expect any breakthroughs in the physical domain to lag significantly - customer service, contracts, sales and coding will be automated, but no self-driving cars and humanoid robots*, and the humans that were formerly in those jobs will be pushed into somewhat less cushy replacements that make use of their skills but also involve some hard-to-automate real-world component - assembling and maintaining bespoke machinery, driving cars, installing cabling, etc. There is a certain possibility that this correlates closely with the jobs that have already been bullshittified, to an extent that the metrics of success in them are now also bullshit - ChatGPT may be a 100x more productive legal brief writer than the human it replaces, but more and better legal briefs could amount to somewhere between a little more and infinitely less productivity. Meanwhile, the humans freed up by this to do more productive work like driving deliveries may not actually be that great at those jobs, so you get something between a slight improvement and a net negative change to baseline productivity while also having to contend with an overall productivity tax from social upheaval (as large strata of the population curb their consumption due to uncertainty or personal socioeconomic drop).
* Always seemed obvious to me once you take away human conceit. In the former domain, you are fighting to outperform maybe 40000 years of evolution; in the latter, some tens or hundreds of millions.
I think it's at least worth considering the other direction too. "What jobs are not currently worth paying a human to do, so nobody does them?"
One place that IMO would be unsurprising (and is probably happening, if quietly) is ML for trash sorting. It's not really worth paying someone to pick recyclable cans from the trash can (something something minimum wage and homeless people redeeming bottle deposits, but that isn't really at-scale anyway), but it seems like something AI could do without hazard pay.
More options
Context Copy link
Also the physical jobs have been getting automated since the invention of the domesticated ox, the wheel, the lever, the steam engine and the assembly line. So what you are left with is the hardened core of physical jobs that are the hardest and least efficient to automate.
More options
Context Copy link
More options
Context Copy link
No one knows. If the firms don't deliver soon, they may cause economic duress to others. Most big firms have invested heavily, where's your ed at is a good lefty newsletter that has detailed pieces on this exact phenomenon.
In a long enough timeline, we Wil automate away a lot, the short term isn't super hopeful.
A complete replacement is very very hard, self driving cars are 90 percent of the way there which means we both still drive daily and will continue that for a few more decades.
More options
Context Copy link
More options
Context Copy link