You are viewing a single comment's thread from:

RE: I Don't Want to Write for Bots that Don't Consider Content. Response to @dantheman latest voting/curation post.

in #steemit8 years ago

@sykochica Well other than the fact you've just presented a direct threat to my business model. It's a well written and well thought out post.

Honestly, this post kind of hurts my feelings a little @sykochica because it sounds like you're saying that your followers don't matter, what matters is the whales or by proxy their bots.

You aren't writing content for the bots. You're writing content for people, especially your loyal and devoted followers such as myself.

Bots will make it much easier and more likely for you to end up on trending.
You won't have to "fake" a social connection that isn't there or whalebate.
A bot is dispassionate, it doesn't have feelings. It's going to make a business decision based on heuristics and that will be a lot more fair than only getting on trending because you happen to be friends with a whale.

Frankly I applauded @dantheman 's stance because of this. Because it removes bias from something that should not have bias, but is rife with personal bias.

You just keep doing what you do and the bots are constructed properly they will pickup and highlight your content from time to time. But your followers are already highlighting your content and that's your audience, us.

Let those of us building the bots worry about making the bots fair and profitable. You just keep making the very best content... Like you always do!

Sort:  

You raise good points as usual @williambanks.

One thing I want to clarify is that I don't have an issue with the bots themselves, but rather that it's being stated that their algorithm isn't taking the content itself into account. I'm all for curation bots to reflect the owners preferences of topics and even opinions, but I don't see how this can be achieved without some concern for the actual content. I interpret this to mean that things with a high word count, linked sources/users/posts, and the like are the key variables, much like old school SEO. I really don't want to see the equivalent of 'hiding a bunch of keywords' (like the old black text on black background trick.) This wasn't reliable for Google search, because it skewed the scoring, causing people to be commonly recommended pages that had nothing to do with the search parameters. I'm sure these bot algorithms will be updated and strengthened over time on Steemit, just like what happened in SEO.

I just find the current state of "content being moot" to be concerning. The double edge sword right now, in my eyes, is that we build a following that can only organically grow so fast and so far (with the people having only 100-150 meaningful, friend relationships) coupled with the rarely stated necessity to trend for further exposure which really requires a whale upvote to make happen right now.

Again, I still plan to keep on doing what I've been doing and It's my own personal principle that prevents me from writing to a bot algorithm. I am sure there are those that disagree with me on this and that is fine. This is solely something I wanted to get off my chest and state my view.

Now onto followers. I completely agree this is important and honestly how I foresee people (myself included) being able to grow, keep and communicate with their following. However, with the ability to follow and be included in a persons feed, is in a bit of a BOOM state right now. I myself am guilty of this if for no other reason that I feel bad when I see that someone is following me and I'm not returning the favor (specifically being unprompted, not the follow for follow agreement.) The current marketing trend, whether good or bad, is to get yourself followed, and for many that means following back. The returns from posting in post promotion has dipped since the feeds turned on. I'm not bashing this, it's both the current strategy (with the feed being new and so many users, myself included, first going through the feed for posts) and human nature in some cases (feeling bad for not following back.) My feed quickly become overwhelmed by articles and many just get lost in the fray, which I assume I'm not alone in having happen.

I absolutely agree that things will get to where they need to be, whether it be curation bot algorithms or me (potentially others) weening down on those I'm following to get a better feed. I just find that we are currently in a 'lull' where we are waiting for both to be progressed. This can be disheartening when dozens of organic likes and great conversation aren't considered important enough for a content blind bot to curate. However I am confident that things will get where they need to be.

P.S. I'm sure I'll be over this in a day or two. I just found it concerning from a minnows point of view that doesn't care for idea of spamming whales with .001 SD transfers begging to have a post looked at.

P.P.S. Sorry for the wall of text, lol. It even makes it hard to proof read this, so apologies if there is rambling.

@sykochica Our minds are similar. I've said that before. We both have a tendency to be prolux. It's not a bad thing. It means we have a lot to say and we care enough to try and build a compelling argument instead of simply making assertions without a lot of thought.

The core of the problem we are facing with this whole bot thing is objectivity vs subjectivity. You can only have subjectivity if you "feel" something, and bots can't be made to feel anything (caveat: I have a way to do that, but it's a topic for researchers not steemit).

Anyways, curation is a subjective task, but there is so much to curate that you really do need a bot. Look at what you said about your feed filling up. And you're right you can't just expect folks to follow you if you don't follow them back. Not until you're prominent and neither you nor I are prominent yet.

So what you really want is a "smart feed". Something that can examine the kinds of things you normally like objectively and then make suggestions based on those criteria.
This is a task a bot CAN do and do well. It's exactly how netflix's "recommendations" engine works.

You don't just "want" a bot, you need a bot in order to give your time and attention to the very best of what your friends are putting out. Not everyone of us is going to hit a grandslam on every post and no one is going to produce content you are 100% interested in 100% of the time.

The ONLY difference between you and a whale in this regard is that the whale has massive influence and needs help deciding not only what to view, but what to upvote.

So the solution to this is to build bots for whales which use objective criteria which match the objective criteria the whales are using. (topic, voicing, and the SEO stuff you mentioned which really boils down to presentation)

And by the way @dantheman was wrong. Bots can examine content and style and make determinations about quality. It's been a solved problem in academia for years now. Just very few of us have the resources available to build and run something like that. Remember Isaac.Asimov ? That was actually a real thing he was doing and readability has a HUGE impact in how your posting is received.

But we can build bots that go WAY deeper than that. Just keep making content for your friends and remember that one of your friends is actually building these things and would never do anything to reduce your exposure. You're a great example of why I WANT to build these things!

The concept of a smart/filterable feed (which I first saw on @mor blog) would absolutely be a big help for Steemians of any scale, minnow to mega whale, for anything from personal interactions (posting, commenting, etc.) to curating.

In the US military bots are used for a wide variety of tasks, sometimes requiring some kind of munition (gun, missle, etc.) While these armed bots quickly handle the massive amount of information needed to navigate, detect and avoid potential collisions, there is ALWAYS a human in the loop that pulls the trigger.

Why?

Because they aren't comfortable giving the firing control over to something not human. There feeling is that the person that pulls that trigger NEEDS to be able to FEEL. Now I understand that the firing controls scenario is an extreme case, especially in comparison to a curators upvote.

This is where I think our views might differ @williambanks. To me by passing the act of curation entirely to a bot, even if designed by the curator themself, that the trigger is being pulled by something that doesn't FEEL.

(I understand that others may not feel the same way on this, it is just my personal view. Honestly I'd be curious how other felt on this.)

A compromise that I would feel much more comfortable with would be a bot filtering through and recommending a large number of posts (i.e. Netflix Movies), in various recommended categories (steemit, politics, etc..), from which the curator then chose the final selections. I would be much happier having a human, that does FEEL, in the loop.

[As an afterthought, the white paper stated a major goal to enhance conversation. With this in mind, there would be great benefit of either weighting comments heavier in the 'Hot' algorithm or adding a comments/time equivalent filter. While votes are important, typically the gold is in the comments.]

Coin Marketplace

STEEM 0.18
TRX 0.13
JST 0.028
BTC 56934.21
ETH 3091.02
USDT 1.00
SBD 2.38