This, however, isn’t shrinkflation. This is supply chain, demand, and uncertainty.
This has been happening in the USA way before 2001:
> In 1967, the Fair Packaging and Labeling Act (FPLA) was enacted to ensure that consumers had enough information to make an informed choice between competing products. The Act requires each package of household "consumer commodities" included in the FPLA's coverage to have a label that includes the net quantity of contents in terms of weight, measure, or numerical count (measurement must be in both metric and inch/pound units).[10]
* https://www.stlouisfed.org/publications/page-one-economics/2...
* https://en.wikipedia.org/wiki/Fair_Packaging_and_Labeling_Ac...
> In fact, it was the humorist Art Buchwald who was among the first to sound the alarm. In a column entitled “Packaged Inflation” published in 1969, he lampooned the growing tendency to conceal price increases. Tongue in cheek, he praised American industry for “devising new methods to make the product smaller while making the package larger.”
[…]
> In late summer of 1974, for example, Woolworth’s offered a packet of pencils at its back-to-school sale for 99 cents – same price as the previous year. But as a sharp-eyed reporter at The New York Times observed, the packages only contained 24 pencils, six fewer than the previous year. The same strategy affected packets of construction paper (24 sheets, not 30).
* https://mikesmoneytalks.ca/shrinkflation-is-an-economic-mons...
And in recorded history for centuries:
* https://archive.ph/https://slate.com/news-and-politics/2022/...
It most certainly is shrinkflation. It's rising input costs decreasing the product quality.
A "good salary" (or at least the median) used to be $5000:
* https://en.wikipedia.org/wiki/File:Median_personal_income_af...
* https://en.wikipedia.org/wiki/Personal_income_in_the_United_...
As the 'dollar has lost value' people have demanded more dollars in their salary. (Whether the two have kept up with each other is a matter of debate/concern.)
The same happens when someone mentions something about the Average income as opposed to the median income. The average income is meaningless if the top 1% keep going up while the rest of the system stays the same. The average would look like people are earning more money when they're not.
Same thing with Market economics and the price of items. If I got a choice to sell a boner pill for $1 to a million people and $1 million to one person, those are equal value propositions. So whenever a corporation raises prices, they don't care how many customers they cut off as long as the remaining ones are whales. That's particularly sharp with all the AI costs being shoved into the pipe.
So, that is to say, you really shouldn't just produce a single number about anything and treat it as some benchmark across the world.
Note: some single item shocks can lead to broad inflation (eg: oil) but that effect takes awhile to play out.
* Google's upcoming folding phone is going to have less RAM than the current model.
* Motorola has both increased the price on their Razr flip phone and downsized the minimum storage
* Sony reduced storage on the PS5 Slim
...
I know Apple is escaping it due to their large contracts but I’m honestly not sure how at this point. They must have pre-purchased multiple years of memory or otherwise have a really insane contract.
But what’s puzzling about that is, why don’t other manufacturers have the same kind of deals? It’s not like Lenovo is a low volume supplier.
Obviously, the iPhone sells in volume unmatched by other devices. But still…I’d have to ask why other high-volume brands like Samsung have wildly expensive laptops.
It just seems like the other companies are asleep at the wheel and don’t have any passion for their strategy, to the point where a tiny company like Framework is overperforming just by caring a little bit. Sure, they can’t beat Apple on raw value but they at least they put together a laptop with a respectable trackpad and a CNC body. Where is volume leader Lenovo?
It's not like designing your own part is free, but Qualcomm charges a very healthy margin on top of their manufacturing costs.
Apple also invests in designing the tooling and processes used on their manufacturing lines.
For instance, they cut way back on how much CNC time was required to produce the Neo.
If lenovo is buying a billion chips a year, why can’t they lock in like Apple?
You can't lock in prices forever, though. The more volume and stability you have, the more a vendor will be willing to enter long-term agreements with you. Lenovo has less volume than Apple and is not in as great of a financial position, so they don't have as much leverage.
The bigger factor is that Apple already had more margin in their products. The price premium for RAM upgrades on Apple laptops is large, as everyone knows. They could absorb more RAM price increases without being forced to raise retail prices.
In the consumer space, I recently bought a Sonicare toothbrush, and the number of models and combinations is staggering. 1000x plaque removal, 750% plaque removal, it's ridiculous.
Lenovo controls less of the stack than Apple: CPUs (Intel/AMD), BIOS, operating system, etc. While ostensibly Apple and Lenovo are both selling personal computers, Lenovo is in a (sub-)segment of the market that is commoditized with Dell, HP, etc.
If you need to run Windows and associated (Windows-only) apps, what's special between Lenovo/Dell EMC/HP/etc? How much of a difference is there between Coke and Pepsi?
A lot of vendors hitched their wagon to the Wintel duopoly, and now they're all riding (or being ridden) herd.
More importantly, if you have a locked in price you can sell your products for more profit - or you can sell the things that you have locked in and not have to make the rest of the widget at all. Sometimes someone will give you a good deal to buy out your locked in contract.
They do this for their own reasons but it's helping them in this crisis. They can simply accept lower margins in the short term, in the knowledge that in the long term these price fluctuations even out.
From the perspective of the producers Apple are a consistent purchaser with deep pockets. AI companies may be willing to pay more for RAM in the short term, but Apple is a safer customer. The current AI bubble may or may not burst, but people will keep buying iPhones regardless. The producers do not want to freeze Apple out because Apple is their hedge against the bubble bursting.
Lenovo may not be a low volume purchaser but they are not at Apple's scale nor are Lenovo's customers willing to pay the premium that Apple's customers are.
- 4TB Samsung Pro 990 SSD for $150 (now $940)
- 64GB DDR kit for a laptop $180 (now $700)
- 64GB DDR5 CL30 kit for a desktop $200 (now $950)
- 9800X3D/5070Ti PC $1800
- 2TB Samsung Pro 990 $95 I think? I honestly don't even remember why I bought this
It's really depressing now. Normally at this point in the NViida product cycle we'd be expected a 50x0 Super series. I think it's all but confirmed we won't see those until next year. I think the 50x0 series will last a lot longer than the 40x0 series.
So it's going to be interesting to see what happens when this hits phone makers who also need RAM. There certainly won't be a RAM increase this year and there'll likely be a price bump. Apple may be able to absorb this to some degree because of anyone I expect them to have long term contracts.
Still, Apple has temporarily delisted the base 16GB Mac Mini and removed the 512GB Mac Studio so they aren't unaffected.
But I think this SSD/RAM price hike has basically killed the Steam Machine, which is sad. Valve obviously didn't lock in long-term contracts before announcing it. Woops. The Steam Deck is also a hard find as a result.
We've seen an almost unprecedented price hike on the PS5, which is an almost 6 year old console at this point. I wouldn't exxpect a PS6 before 2028 at the earliest.
We've had RAM price spikes before, usually because of supply crunches (eg years ago I seem to remember a fire taking out one of the major suppliers).
I honestly don't expect any of this to get better until we have an increasingly likely global recession and the AI bubble pops. OpenAI and Anthropic may not be able to cash out in time to avoid all this.
If seemingly the "same" product -- in this example say 'MacBook Pro' base model for current year -- delivers less goodness, less value compared to its price year-on-year. If the price appears to stay more or less the same but the product is made weaker in service of higher margins for the seller. THAT would typically qualify as shrinkflation ... in my understanding.
This is more objectively measurable in comsumer goods where you can see the packaging being tinkered with over time so consumer thinks they are getting the same SKU but this year's packaging has less of the product tghan last year's, at similar price point so it doesnt register as price inflation.
We just changed that minimum to 10TB last week and it was specifically the RAM prices that forced us to re-calculate that value proposition.
This is the same product at 5x the price. It sucks but it's not shrinkflation
ASIC LLM?
inference inside SSD?
bigger 1-bit trained models?
better MoE so we don't have to load all experts?
Alternatively "AI" gets such a bad name regular people will go and burn it all down as long as it has "AI" even just in the name.
Many product segments peaked and the only way left to extract more money from us is to either lower the quality so that it's cheaper to make/break faster or subscriptions/ads.
Maybe it seems like this strategy leads to a permanent decline in global labor power, but history shows a different pattern: "where capital goes, conflict goes". Relocating capital to exploit cheap labor does not permanently resolve crises of profitability; it merely reschedules them in time and space. By moving to new regions, multinational capital inevitably creates and strengthens entirely new industrial working classes in those areas.
Conversely (complementarily) when an industry becomes too crowded and profits are squeezed, capitalists do not just cut corners; they rely on what Beverly Silver terms the product fix—shifting capital entirely out of mature, highly competitive sectors into new, innovative, and more profitable industries.
Historically, the epicenter of capitalist accumulation (and subsequent labor unrest) shifted from textiles in the 19th century to automobiles in the 20th century. In the first decades of this century, capital shifted toward semiconductors, the "education industry," and producer services (like finance, telecommunications, and consulting).
Because a product fix involves withdrawing capital from an established industry, it usually brings about mass layoffs, deindustrialization, and the breaking of existing social compacts. In response, the workers who previously benefited from those compacts have, historically, risen up to protect their jobs, pensions, and established ways of life.
Unfortunately, they are often doomed by their diminishing economic leverage.
Why do we always act like there's an immutable social obligation to march right along believing the prior generations had freedom to start marching in that direction, but we are forever locked in to such a direction.
You know all the people that made those choices are dying and future generations have zero obligation to carry on linearly from where they left off?
Women would not have the right to vote. We'd all be speaking Latin.
Two things that would remain true if society of the living was tightly coupled to exactly how the past worked.
We actually don't need all the ram. Everyone was fine in 2010. The devices were fine, the internet was productive.
We have all just seriously fucked up in the software and hardware space. We are super-sizing devices and software just as surely as the car industry has done in massive pickups and SUVs and the food industry has done with portion sizing.
Oh, the irony
So, we can't ship. This could kill our business. Thanks, Sam.