Unfortunately that visualization doesn't really work with vector dbs. Vector dbs normally split their data into separate segments and build indexes on them separately. There is no one overarching index but rather many small ones that are searched in parallel. In addition to this, such a high compression to reduce it down to 2d/3d ends up becoming a giant blob without too much information.
Maybe I'm misunderstanding what exactly a vector db is.
Let's say you have a chatbot and stored in its database is the usual info like session id, timestamp, message etc. To "vectorize" this db then would be to vectorize all the messages. Is this too simple an understanding?
Once the db has been vectorized then we can do semantic search on the messages and create more informative graphs based off the semantic similarity for messages within a given timeframe or other criteria.
It lets you run you run the benchmarks using your own API keys. Although it is made by Zilliz (maintainers of Milvus), you can take a look and see what is going on and judge if its fair.
The article isn't meant to present a new library competing with Annoy or other state of the art approaches, but instead, describe how vector search works using short code snippets.
I've now updated the article to clearly mention this and also cite annoy.
The description of this is kind of confusing but I think the easiest way to understand it is that it is a data processing pipeline of sorts. Take unstructured data and apply transformation and computation. A similar project to this is Towhee (https://github.com/towhee-io/towhee). This project tries to simplify unstructured data processing and provides pretrained models and pipelines from their hub.
Usually this is done in three steps. The first step is using a neural network to create a bounding box around the object, then generating vector embeddings of the object, and then using similarity search on vector embeddings.
The first step is accomplished by training a detection model to generate the bounding box around your object, this can usually be done by finetuning an already trained detection model. For this step the data you would need is all the images of the object you have with a bounding box created around it, the version of the object doesnt matter here.
The second step involves using a generalized image classification model thats been pretrained on generalized data (VGG, etc.) and a vector search engine/vector database. You would start by using the image classification model to generate vector embeddings (https://frankzliu.com/blog/understanding-neural-network-embe...) of all the different versions of the object. The more ground truth images you have, the better, but it doesn't require the same amount as training a classifier model. Once you have your versions of the object as embeddings, you would store them in a vector database (for example Milvus: https://github.com/milvus-io/milvus).
Now whenever you want to detect the object in an image you can run the image through the detection model to find the object in the image, then run the sliced out image of the object through the vector embedding model. With this vector embedding you can then perform a search in the vector database, and the closest results will most likely be the version of the object.
I agree about it being unprofessional. But does "virtue signalling" now mean simply any form of protest?
Were the Canadian truckers protesting in Ottawa virtue signalling? Are people posting "Let's Go Brandon" online also virtue signalling? If not, what's the difference between their form of protest and this Terraform stunt?
My understanding is that "virtue signaling" implies that the primary goal is performative with minimal personal risk and minimal commitment to productive action. The example that comes to mind is a company that spends far more money informing the public of their charitable works than they do on the works themselves.
So an in-person march, trucker protest, sit-ins, having a private conversation, calling your representative, making personal sacrifices, attempting to bring attention to lesser-known issues, donating money, engaging in dialog to convince someone of your position, etc would not be virtue signaling.
But things like posting "Let's Go Brandon" online, or changing your profile picture with no further action, or unironically using terms like "virtue signaling" for internet points might qualify as virtue signaling.
> My understanding is that "virtue signaling" implies that the primary goal is performative with minimal personal risk and minimal commitment to productive action.
And it's based on the flawed assumption that stating public support for something without doing anything else is useless.
But people moderate their behaviour based on perceived social norms.
When people publicly state their support for a given issue, they are communicating what they understand social norms to be.
When a lot of people do that, that becomes the norm.
So "virtue signalling" could just as easily be labelled "showing support", which is the way that we share and align on those norms.
But, of course, folks who don't like people voicing their support for those values, for fear that they will become normalized, needed to find a label to apply to insult those people and, hopefully, stop people from voicing their support for these social movements.
And thus the term "virtue signalling" was born. Suddenly saying out loud what you believe becomes itself a social moray.
Now flying a pride flag, or calling for increased diversity in the workplace, has become "virtue signalling" and something to be embarrassed about.
It's quite clever as a means of controlling the narrative. And it appears a shocking number of people have bought into the BS.
1) Does "showing support," actually do anything? Are we really aligning on norms or just scoring points with people who already agree the same position? I suspect the detail matter and that there is continuum, where for uncommon positions maybe it does something, but for widely held views, it really is just "virtue signalling."
2) When does "showing support," become a substitute for more substantive action. Maybe I post a pride flag on my social media avatar, but don't bother to vote in a local election with discriminatory ballot initiative. Or consider any number of incidents of corporate "greenwashing."
But sure, plenty of virtue signalling, isn't _just_ signalling. And we shouldn't dismiss it on those terms, but rather ask about impact.
How much is a heap of sand? It's a Sorites problem. There exists a continuum between "changing your profile picture to be tinted like flag X" and "everyone is saying and doing the same thing and it's an entrenched social norm".
It's certainly less impactful than doing something substantive but it also costs nothing to signal boost. Same like boycotts, it only works en masse.
I think one of the important things about virtue signalling is that you're making a big deal about the prescribed norms that everyone is, in our unofficially official ideology, supposed to follow.
It'd be like flying a great big flag that says "I support the government and corporations." Whenever I see a pride flag, that is essentially what I see. It's like, (and it's useful to read the people who are against you as a mirror), when Patriarch Kirill of Russia says: ". Today there is such a test for the loyalty of this government, a kind of pass to that “happy” world, the world of excess consumption, the world of visible “freedom”. Do you know what this test is? The test is very simple and at the same time terrible - this is a gay parade. The demands on many to hold a gay parade are a test of loyalty to that very powerful world; and we know that if people or countries reject these demands, then they do not enter into that world, they become strangers to it."[1]
That's not far off from the truth. Virtue signalling, as opposed to protest, is oriented towards moving closer to power, rather than further away from it. In Foucauldian terms, protest would be a transgression, a breaking of the taboo, and virtue signalling would be the opposite of that, an adherence to and reinforcement of the taboo, in which one mimetically serves the strengthening of the taboo, until the mimetic crisis breaks into blodshed.
Virtue signalling is wearing the swastika in 1938, and bears no relation to wearing one in 1929, except to say that those who wore it in 1929 won.
At least here in the United States, those are usually sold by charities that benefit various veteran's organizations, so there's some actual skin in the game there.
Good question. I think it would on the motivation of the wearer.
If if they are wearing it primarily as a method of socially fitting in and signaling to those around them then it might be "virtue signaling".
If they would continue to wear a poppy on remembrance day in contexts where the people around them did not know what it meant then there's clearly something more than signaling going on. Perhaps it is their tradition, or it helps them remember.
(I should caveat that I don't think there's anything particularly wrong with social signaling so I don't use the term "virtue signaling" as it seems to have a pejorative or sarcastic connotation.)
Just about all of your examples of not virtue signaling could or could not be depending on the context. Going in person to march, posting selfies on facebook, marching two blocks and leaving would be an in-person march and virtue signaling for example. Virtue signaling is more about intent and advertising of the act.
To paraphrase the bible "Jesus said The pious pray in their closet. Those who make big shows of praying in public are nothing but douchebags."
You just successfully moved the goalposts by redefining what "going to a march" means...and even then didn't address how "all the examples" don't count. Even then, showing up for a mere two blocks still involves more risk than staying at home.
To me, it's the element of risk that differentiates virtue signaling from meaningful action. Posting "Let's go Brandon" on parlour or "Black Lives Matter" or Tumblr aren't risky actions. Saying "I think gay marriage is ok" in a conservative church is. Just because you can imagine a situation where the context lessens the impact of the action, doesn't mean the example is weak or wrong.
Then you missed the point of my post entirely. "Virtue Signaling" isn't about what good deeds you do or don't do. It is about intent and what you do "around" those deeds. You can do real good and still be virtue signaling, because virtue signaling is the advertising aspect of it. Two people do the same deed, one won't shut up about it. One is virtue signaling, the other is just a good person.
The goals/motivations of the action is what matters here. Changing a few lines of code to state your stance on an issue won't cause any level of change whatsoever. It was clearly done to show which side of the war the author supported, more about the author than the conflict. A Canadian protestor that spent most of the time publishing their involvement on social media is virtue signaling, but one that merely occupied the capital is not. Saying "Let's go Brandon" is virtue signaling, unless the signal is meant only for your group, then it is dog whistling
If you support a cause, "virtue signal" describes an action which doesn't do a lot to materially support the cause, to encourage people to take more concrete action.
If you oppose a cause, "virtue signal" is a term of denigration for any public action on behalf of the cause to discourage people showing support for it.
My personal sense of the word is that “virtue signaling” is when people intentionally seek recognition from a group by visibly supporting something that group already endorses or considers normal. Going a step further, the support is often exaggerated, not totally sincere, or not congruent with that person’s previous behavior.
There is also a sense that whatever thing someone is “virtue signaling” about is acceptable enough that there is no real downside for taking the stance.
It would be like an American proudly declaring how much he loves the United States on Independence Day. He would go out of his way to emphasize just how much of a patriot he is, hoping to be rewarded for doing so.
I believe “virtue signalling” means any form of protest which will not, and does not really try to, have any actual effect on that which it protests against. Any small actor protesting and boycotting something which they are unlikely to affect and even come into contact with, therefore qualifies. The protest is not done to affect any real change, only to signal virtue.
When you protest something not because you care, but in order to signal that you care it is virtual signal (literally, you are trying to signal your virtue).
When you protest for gay rights in 2020 it is a virtue signal. When you protest for gay rights in 1987, it is because you believe in it and are willing to take the cost of it.
Since public trade companies only care about money, when those companies support some course it has become safe enough that it is now virtue signalling.
> When you protest for gay rights in 2020 it is a virtue signal.
I don’t think this is the case; it’s still legal to discriminate on the basis of sexuality in many states and contexts such as housing in the United States. You may not be able to fire someone for being gay but in many places you can evict them for it.
Additionally I don’t think it’s virtue signaling to protest for gay rights beyond the small number of countries that recognize marriage. There are still countries where gay behavior is illegal, marriage isn’t recognized, and gay panic is a legal defense.
Like all politically-charged terms, the original meaning has been long lost as the context in which it was coined is forgotten.
I think the term ‘virtue signalling’ was originally intended to point out a perceived hypocrisy - that it’s much easier to gain public support for an idea/cause/campaign if that campaign is perceived to be helping some disenfranchised group - even if the campaign also benefits the organizer, and even if the campaign is not necessarily wanted by some or even all of the allegedly-aggrieved group.
It did start out that way (an attack by conservatives against progressives) but has since become ubiquitous. It just refers to immaterial forms of protest that don't accomplish anything except signaling support for a cause.
I think people virtue signal when they dominate conversation with pet political topics. This is especially evident if they continually find non-sequitur ways to include moral and pet topics in regular conversation. I don't think that either of the examples you've brought up are virtue signaling, but introducing the topic of Starbucks cups to redirect a conversation into how "the country has lost its way" is a good example. An example of this on the left is how certain folks will redirect any conversation into one about oppression.
The impact of virtue signalling is pretty evident. Ever seen how in order to make a statement on something you have to first identify yourself as part of that something that you're criticising? That's a direct byproduct of virtue signaling.
Accusing people of virtue signalling is a prime example of virtue-signaling.
But really I don't know your motivations and you don't know the motivations of the Terraform folks. Let's be a bit more humble please. But yes, I also find it unprofessional.
