SAFE Network Data Inflation

Continuing the discussion from Dealing with copied content:

To me, this seems like a very significant risk for the economic model of the SAFE Network—so I was wondering if others agree, and if it has been seriously considered. If there is a fundamental incentive in SAFE’s design to inflate one’s data, this would seem to lead to significant long-term inefficiencies and problems.

Yes, there could be factors to limit more extreme data inflation, as janitor pointed out:

However, more subtle inefficiencies could add up over time, especially if it became a systemic issue for SAFE.

Possible examples include:

  • Choosing less efficient lossless compression over more efficient lossless compression.
  • Encoding files at unnecessarily high fidelity (e.g. an audio file at 24-bit/192 kHz when the audio in question does not significantly benefit from it).
  • Adding superfluous images to an otherwise text-based book (even a single large title page image, or “About the Author” image, could significantly change the size of a relatively small book).
  • Or simply inflating the content itself (e.g. making a film or piece of music slightly longer, regardless of artistic considerations).

If the inflation is not particularly significant or noticeable for most consumers, there will be limited incentive to seek out ‘reduced-size’ versions that are subsequently uploaded by others. On an individual basis, such issues may seem minor. But, as an aggregate effect on SAFE’s efficiency over time (especially on the scale of millions of unique files), it seems like it could be quite important.

There will be people who do this, but whether it is an issue or not is hard to say.

Firstly, there’s no incentive to bloat anything unless it is being accessed. So everyone wanting to make money from publishing content will have to first focus on making it popular, which means useful, and they will be in competition with others doing similar things.

I think this establishes a quality first mindset, I think it also makes it hard (though I’m sure some will) for someone who has achieved popularity, to risk losing it by a) shifting their focus from quality, usefulness and virality, to puffing up the content, and b) causing their content to lose appeal relative to others by taking longer to load.

I’m not saying it won’t happen, but I think it is likely not a great strategy, and therefore not clear that this is a significant threat to the economic model of SAFE network network.

In extreme cases (such as the previously given example of converting a book from text to images), and in cases where the bloat is easily apparent (like websites with unnecessarily heavy use of audio-visual media), I think you’re probably right.

In more subtle cases, however, (such as adding a small amount of bloat here and there—likely under the guise of increased quality or added content) I’m afraid it would be more difficult to distinguish. And, as I suggested before, it seems possible that bloat which is seemingly insignificant on an individual level (and thus likely not large enough to affect users’ consumption patterns) could be much more consequential on a network scale.

Of course, I hope you’re correct in not worrying too much about this. I want the SAFE Network to succeed, so I hope any failures on this point will be limited to failures of my own understanding, and not of anything inherent to the Network itself.

1 Like

Is the reward the creator only paid on the first chunk of the file? Or is it paid on every chuck that is GET? By paid I mean using the formulas&lottery

Until this question is answered can the discussion have full meaning. If only paid on first chunk then bloat is meaningless to payments, if not then bloat is and invites others to repost an unbloated version.

[EDIT] Answered - any chunk

Any get for any chunk (immutable data).

[edit] I should add this will likely be only for public data. That is an easy remedy and seems very reasonable. Some thinking is along those lines and I feel it makes sense.


Thanks David

So now we will see people unbloating content and reposting it. Eventually those who deliberately bloat will see rewards initially then very little as others unbloat it and people learn to look for unbloats.

1 Like

Yes logic should prevail and be in line with market forces here I think.

Doesn’t really make sense to reward the creator for private data does it?

No I cannot see where it would be good to do so, it’s a very strong argument we only reward public data. It’s a +1 from me for only public data, plus less work for the network. It also allows us to add in further logic to make sure it’s widely read first. That could lead to much of the I can create rubbish and read it to game the system. Sad we would need to, but it is pretty easy. Probably as we see results we would, but I feel as the network grows the scammers will end up depleting many more resources than they would ever gain. We will measure though.

New world I hope :smiley:


That is my thoughts too (until its live and can be measured.)

Sounds like more thought has been done since you last talked of Reward The Creator (RTC?)

To reward private data (even if shared) would really only either gain the person a tiny amount or be for the gamer trying to read his/her data continuously trying to earn those one or two SAFEcoins per year. That is trying, but unlikely to succeed

1 Like

But, even when rewards are limited to public data, doesn’t the risk still remain? Would the network still remain balanced and efficient when the data inflation (on an individual file scale) is minor or difficult to discern—but is potentially a network-wide business practice?

Consider the following example:
A well known pop singer with a large Internet following has an ‘official’ website on the SAFE Network. She regularly offers public downloads of her music there. In publishing a file, she decides to inflate the size 10% over what is necessary. (Such inflation might be done in a subtle way, such as re-encoding the audio file at a fidelity higher than the fidelity of the equipment used to produce it.) The fact that such inflation even exists, or was a motivating factor, may not be apparent to users. Regardless, another person (“User X”) suspects that the file was inflated, and re-encodes the file at a lower bitrate/resolution (reducing the file size by some number of MBs). User X uploads the reduced file on another website and offers it to the public.

How likely are users to download the file from the ‘unofficial’ website instead of the ‘official’ one—especially when the difference is so minor and the existence of inflation is only speculative? Would minor, non-obvious bloat be enough to affect most consumers’ download habits? Could the effects of such inflation on an aggregate, network scale not be of significant consequence to the balance and efficiency of the network?

Again, I ask this out of concern for the well-being of the SAFE Network. I’m hopeful that others may be able to explain why such concerns may be unfounded.

I think this is key, I feel as humans we are less tolerant of corruption and folk doing this would be called out and others would reduce the size and republish. All in all it’s very possibly a crazy way to go, maybe akin to adding more ads on a website really. Worse user experience loses users.

I suspect anyway :slight_smile:


How would that would matter to the network?

From an engineering standpoint, I suppose it might not pose any direct problem for the network’s ability to function. But, if data inflation became pervasive enough, I wonder if it would negatively impact people’s perceptions of the network, or spur competing networks. And, perhaps there’s some intuitive bias on my part—but it does seem like if a system incentivizes inefficiency to a significant enough degree, it risks long-term sustainability. …Or, maybe as I feel the network getting closer to launching, I’m simply finding needless things to worry about.

I imagine it is the human aspect of the network that is the most difficult to predict. That being the case, I hope you are correct that the SAFE Network and people’s interactions with it will balance each other well, in a mutually beneficial way. I guess we will all just have to wait and see.

Hi all, first time out, so correct me if I’m wrong…

I’m under the impression that it costs resource to use resource. To store a bloated file would require you to either pay our provide enough space to do so. The bloat is self managing there - there will always be space.

As for discouraging bloat… well, you start publishing unwieldy or ridiculous content and I will personally take it, optimise it and provide it back to the world with no regards for my own income, or yours. Because your stuff is there for me to have and you have proven yourself untrustworthy.

We’re looking for the perfect state of electronic anarchy (if we can put the cultural flavour of the word aside). In that world, you take the Michael and people will write you out.

I can’t tell you how excited I am to be here, by the way.


If the bloating is ridiculous (2x, 3x), someone will repost a more efficient version of the file.
If the bloating is tiny (e.g. 10%), it won’t be noticed by most users (but others may still repost an optimized version of the file).

1 Like

My thought is that “if it would negatively impact people’s perceptions of the network,”, then that means it is noticed and as others said, people according to human nature will correct this imbalance by uploading un-bloat.

But seriously if we try to have a perfect system that stops 10% bloat from occurring then I guess we would be waiting years for all the other areas to be perfect too. Maybe each user has a overseer that ensures perfection. :stuck_out_tongue: Seriously small bloat will occur as people test what they can get away with, but I would think in general that creators will have enough to do without trying to bloat just enough not to be noticed.