Storj buys an AI company

[asturking]

Is this the new big customer that is going to fill up our hard drives and erase data every 30 days?
I’m warming up my router

[jammerdan]

@asturking

Is this the new big customer that is going to fill up our hard drives and erase data every 30 days?

Storj acquiring Valdi

It is not. The company won’t be announcing new customers or customer deals with the Community in general. There may always be reasonable announcements when it makes sense for the business, but in general customer deals are kept internal for a variety of reasons I won’t list here.

[jammerdan]

@Alexey

But when they are talking about a company they become a very toxic, I’m sorry.

But sometimes entertaining too. I don’t know what else to say. Maybe he bought some STORJ when it was at ath?

[Alexey]

I do not think so. I think (and maybe I’m wrong), they just like to be a troll.
And a good one, they managed to involve me (shame on me, honestly… because I know - never feed a troll, especially if they are selected a green avatar…)… oh, but well. We are here now.
But I would expect that they would continue to train me, when they are return from the forcedly vacation. I do not think that they would have anything useful to say about the company, but maybe they would be useful to help to configure nodes? Who knows…

[Eugene]

@Knowledge

This is inaccurate.

Could you please support your words with some statistics or other information?

IsThisOn is raising a valid concern about reality of real customer’s data.

It can be observed through the sequence of the following events:

1. Sacrifice. Test satellites were removed some time ago (minus test data). My rough estimate was around 5% of data stored on my nodes, but had only a few fairly new nodes at that moment.
2. Nothing Personal. Announcement to remove free 25GB tier (minus real, but unpaid data). Can’t estimate what’s the percentage of data on my nodes was removed due to this event. But at this point in time I truly believed that most the data on my nodes is real customer’s paid data. What would I know…
3. The Great Dump. All of a sudden around 2/3 (or 66%) of my nodes’ data were dumped to trash over a short period of time. Older nodes lost more data, newer nodes lost less data, but on average around 2/3. (minus test/real data - pick one)
4. Rebirth of The Phoenix. Over the course of 2 months the data were restored to almost the same levels as before The Great Dump and it was done at unprecedented bandwidth speed. (plus test/real data - pick one).

Honestly, now I think that SNOs are like Guinea pigs (test department) for Storj. Looking from this perspective at all of the past questionable decisions, events and statements explains them in a way, so they start make sense.

[Knowledge]

Test data is on Salt Lake. The rest is customer data. Some customer data was part of the free tier which was often abused by users using Chia Plots, primarily in the EU. The resr is customer data.

Which has been explained numerous times. What customers do with their data is up to them. They may add, delete, at their leisure.

[Alexey]

Storagenodes will always store and delete data, it’s their purpose. We would improve the product, would work on integrations with GPU computing to be part of the Storj Ecosystem more closely than before. And of course we could introduce bugs and fix them as quick as possible, and tests would be performed almost constantly by Storj and by the new customers (they firstly want to make sure that it suits their needs). But most of the not test usage from the established customers anyway.
However, it’s absolutely doesn’t matter is this data test data or not, since it’s paid.