[froidpink]

Meta's profit has increased almost 2x since 2023. Meta makes money from advertisers spending money on Meta. So the profit growth from Meta does very much come from the real economy

[bee_rider]

It comes from the real economy in the sense that it is money they could be spent on something productive, but was instead wasted on ads.

[piva00]

In a video I watched recently there was a breakdown of how much a plumbing company had to spend on "marketing" (aka: Google ads placement, Facebook/Instagram) to attract customers and their per-click pay was about 60 USD, they were spending around 16-18k USD per month on online ads to keep the business afloat.

I had no idea that physical small businesses like that needed to spend so much on marketing just to be found.

[jdross]

The very first thing you discover as a business owner managing cashflow is how hard you must work to get customers versus to provide a service.

[giancarlostoro]

Its worse if you get bombarded with negative reviews, which is why stuff like yelp holding you hostage is so bad. If you own an independent business outside of marketing, your reputation is everything. Especially now that anyone can blast your name because they didnt like how you said shiboleth or whatever.

So if you have a zillion negative yelp reviews, which you have no idea where they came from, since there's more negative reviews than you've ever had customers, but they want your money to hide them. ;)

[ch4s3]

Have to seems like a strong phrase. I found my last plumber and window guy on a facebook neighborhood group. Local ads for general services can be quite expensive, but doing some marketing through local groups only costs your time. I can see how driving business though clicks is attractive, but I'd be surprised if there was no alternative.

[piva00]

Have to if you want to scale the business, if you are a sole trader doing small gigs it's probably very achievable to only use local groups. If you rely on people searching for "plumbers in <X> city" while running a small business with some 5-10 folks I don't think you'd get enough work.

[bryanlarsen]

You don't want to use a plumber that has scaled their business. That means that they're sending out a new hire to do your plumbing rather than the plumber that originally built the company's reputation.

The best plumbers spend $0 on advertising. They've got enough business through repeat customers and word of mouth to keep their small set of plumbers busy, and they're expanding slowly enough to properly train apprentices and ensure quality.

[ncruces]

And, as a customer, you find that plumber how? What about when they tell you they'll be available 2 months from now for a 1 day service?

[insane_dreamer]

That's great for already establish plumbers, not for new ones.

[wiether]

> In a video I watched recently

Macy with Morning Brew ?

https://www.youtube.com/watch?v=ucWsaVbEu78

[piva00]

Exactly that one, couldn't remember at the time I posted, thanks!

[xnx]

> they were spending around 16-18k USD per month

Unfortunately, this doesn't mean too much without knowing the size of the business.

[msgodel]

Sales are the most important thing in business. Everything else only matters insofar as it drives sales.

[bee_rider]

Businesses aren’t real, they are just imaginary things that people use to help us organize the act of building and distributing things.

[cm2012]

Ads are a necessary matching method in capitalism, there is no better alternative unfortunately.

[bee_rider]

I don’t think this is true, but I also don’t know anybody way to verify it either way. Enthusiast forums were always a better source, although native ads can mess that up.

But, even if ads are a necessary evil, they are definitely overhead (in the sense that they don’t actually accomplish anything, just influence the decision as to what should be done). Maybe we can define some sort of ad-efficiency metric for an economy; what percentage of the money is spent influencing decisions, what percentage is spent actually implementing the decisions…

[cm2012]

Here is a longer version of my thinking: https://news.ycombinator.com/item?id=44714407

[msgodel]

I don't know if that's quite right. A nice business directory would probably be better but most people don't want to settle for that.

[cm2012]

Here is a longer version of my thinking: https://news.ycombinator.com/item?id=44714407

[msgodel]

Nah I don't think the directories need to be reviewed (except for listings for things that don't actually exist.) Just show the listings in lexicographic order like how phone books worked.

We already have the law as the meta norm. Let law enforcement do its job.

[cm2012]

Who is making the directory?

Is it public like Wikipedia? How do you decide which businesses are notable enough?

Is it the government? Of which country?

Can anyone submit without approval? It would get spammed to ruin.

Etc etc.

[blitzar]

Right around the time Meta stopped setting 100's of billions a year of cash on fire in the metaverse and pivoted to Ai.

[dragontamer]

I thought Meta put a total of maybe 30Billion in Metaverse?

A lot of wasted money but not 100B++

[cedws]

Reality Labs is still a thing, they’re still working on VR/MR.

[philipallstar]

To open-sourcing AI models though, no?

[blitzar]

To pay $100mil to ai devs.

I have a theory that Meta execs was so focused on the Metaverse that the Ai team succeeded thanks to the lack of supervision and interference from above - there was probably a board discussion between 2019 and 2022 about firing them all and just focusing up the Metaverse stuff becuase they were dead weight on the core mission of colonizing the Metaverse.

Turns out the Ai team was the lifeboat to save the drowning Metaverse.

[buyucu]

In my experience success of a project is inversely proportional to executive attention. The best thing seniour leaders can do is to get out of the way.

[froidpink]

Meta has been investing in AI for more than a decade

[gary_0]

Longer than that, for sure. They were deploying ML systems at least as far back as 2011.[0]

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

[blitzar]

And yet their quaterly and annual reports dont mention it at all till 2022 and the team are now being "helped out" by new $100mil talent.

There is an amazing team there that did the work, I am just saying it wasnt the visionaries vision that made that happen - and if it was they certainly wouldnt have let the Ai team publish or opensource their work.

[RC_ITR]

Meta is also a great example of AI leading to higher user engagement today.

Reels isn't powered by Transformers per se (likely more of a complex mix of ML techniques), but it is powered by honest-to-goodness SOTA AI/ML running on leading-edge Nvidia GPUs.

I think, because they're so impressive, people assume Transformers = AI/ML, when there's plenty of other hyperscale AI/ML products on the market today.

[arethuza]

What if a lot of that advertising is from AI companies that are likely to fail in any downturn - didn't advertising drop fairly sharply during at the end of the dot-com bubble?