Hacker News new | past | comments | ask | show | jobs | submit login
Data isn’t oil, whatever tech commentators tell you: it’s people’s lives (theguardian.com)
80 points by dredmorbius on May 30, 2021 | hide | past | favorite | 24 comments



Oil is money. Money is time. Time is people's lives. When oil is more expensive ordinary people spend more money, and thus more time, for travel and other goods and services. They spend more of their life time to get less. My own life would have been much smaller and tougher without oil technology in the world.

Both data and oil are about people's lives.


I like how idlewords puts it: its like oil. toxic, sticky, and dangerous. flammable. a liability.

https://idlewords.com/talks/haunted_by_data.htm

honestly I forgot there was a "good" version of data is oil because of that talk


All analogies melt if you push them loudly enough.

Data-as-oil having long-lastinga and difficult-to-distinguish unintended consequences, yes, that fits.

But where oil is very nearly the distillation (so to speak) of a perfect universal commodity fuel with universal applicability, data is a near-completely differentiated good, a power multiplier, with rapidly emergent properites of scale in aggregation.

Oil is useful to anyone with a stove or simple engine. At the individual level, only personal or general envrironmental data is either available or useful to individuals, whilst aggregators can make use of large pools of data to manipulate, coerce, intimidate, bluff, and call blusters, at scale.

To that extent, data and oil are quite different.

with emergent benefits to scale coming only at the very largest levels (most especially through imposing embargos), and serves mostly as a capital multiplier

data is both infinitely differentiated, and has rapidly-emerging properties in aggregation as a power multiplier. Individuals can largely benefit only from their own specific data.


Oil was peoples lives, at first. Oil barrens manipulated whole towns out of their land, stole, lied, and destroyed livelihoods. Oil still destroys peoples lives, but the damage is abstracted.


And like people's lives, no one cares about anything until they get the bad news. Finding out your data has been used against you is like getting news of a major health issue. Most people knew the potential problems, they know what is good and bad for their health but consume regardless. This is no different to our attitude with personal data too, IMO.


Data is like Uranium. Make damn sure you contain it properly. And if it gets out you’ll spend decades dealing with the consequences.


Only if you're the one containing your own uranium supply. There's no environmental regulations yet to punish companies that dump it in the local river.


We have data lakes, why not data rivers too?


Also like uranium in that you don’t want to just collect and store it for no reason or some future potential reason. You should treat data collected as a dangerous liability, not a profitable asset.


Except someone stole your uranium and then accidentally dropped it on your house while trying to sell it for profit


It's more like mako energy from Final Fantasy VII (which, it turns out, was actually people's souls).


Shinra theme song playing in background


This is why we need legal work for data access / use. The legal system is a reflection of us.


Not all (or even most) data is personal.


Most data by itself is useless, even things we consider "private" like SSNs, credit cards, etc. They're like $20 for a bundle on the black market. But if you can link that stuff to browsing, location, or transaction history? Anyone with that can read you like a book. It's not about what data is collected so much as it is how much data is collected - and how much of the collection can be traced back to you through a cookie or IP address. That's the big problem. Anyone with enough random data about you can put together your life piece by piece.


The best defense is to reduce the size of the attack surface by living a life so bland that even perfect knowledge about you is altogether useless.


not really, the real strength is in predictive power. face recognition at one point stored your face as just 13 parameters, maybe our brain can be compressed in 20 billion parameters and reasonably approximated by ml with just 100 million parameters.

if you want to go to the extreme you can say that every bit of data you give to Google/Facebook/Amazon goes toward training this mooch of Human simulation engine (which would be different from general artificial intelligence)


I hope that's sarcasm.


It is indeed.


Even so, there are many people who adopt it as a serious approach to life. Some consciously and some unconsciously.


Yeah, I'm usually good at picking up humor online, but I see so many people wasting their lives trying to appease corporations in ways that ultimately don't matter. It's very sad to watch.


In the context of mainstream news reports on big tech, the "data" they're referring to is personal, or at the very least able to cause behavioural changes on an individual level.

Journalists don't care about non-personal data ("only x% of users use the symbol insertion tool in Word 2016") - even though Microsoft certainly does - because it makes for a mundane news story.


But this shouldn't matter. We still have a subconscious and we sometimes do things that we don't really think through, or do them on auto-pilot. So you could be doing things which you don't think through and they are still indicative of who you are as a person. But now, you don't know this part of you and you end up in a situation where some algorithm running on some servers somewhere is a better predictor for what you want (among other things) than your own thoughts.

Truth is people seem to not care about their subconscious. And they fail to understand the power of data when combined with 1 million other people. I still hear the response 'Well I have nothing to hide' way too much from people living in 2021.


A fair point, though the data of interest to marketers, and others making use of targeted influencing and/or surveillance, is.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: