razharov
Senior Member
- Oct 19, 2022
- 898
- 961
Wait until you can only get your page indexed if you hand submitted it via GCS, after biometric authentication. (PS: I don't really think we are so far away from this)
Google openly denying AI content from now on.
- Thread starter razharov
- Start date
- Mar 21, 2013
- 24,786
- 31,561
Google is getting desperate. GPT is much superior to Google as a search / research tool.
This is a direct attack against OpenAI IMO.
razharov
Senior Member
- Oct 19, 2022
- 898
- 961
Yea, somehow this new google release is "timed" well, according to the new AI relates... I swear google spends 50% of their workers to reverse engineer AI-s and learn to how block them
You guys really think a free to use AI can hold a candle to Google's in house developed AI that has the processing power of multiple supercomputers at his disposition and is coded by the top engineers on the planet?
Sounds like a lot of copium and hopium to me, the fight against google is even worse now.
It used to be bot vs bot and you could compete now it's AI vs AI and you won't be able to compete at all, don't know how people don't see this.
AI increases the difference in power, it doesn't decrease it.
Sounds like a lot of copium and hopium to me, the fight against google is even worse now.
It used to be bot vs bot and you could compete now it's AI vs AI and you won't be able to compete at all, don't know how people don't see this.
AI increases the difference in power, it doesn't decrease it.
Recent AI contents are solid gems. I don’t think Google can detect every single AI content just by checking the single content. They will most likely to check the quantity of the articles published in the period of time and the probability of it can be written by humans.
Exactly people forget or don't understand that AI (to simplify it) is a bruteforce technology. This type of tech works by taking information as input (google has all the information) and then using processing power to get to the desired result (google has all the processing power), also the AI updates itself nonstop(hard to be faster than that). Readable content that makes sense is not necessarily content that a Human would write, it will always be detectable by an AI that has google's infrastructure and power behind it, (it won't trick most humans either).
- Mar 21, 2013
- 24,786
- 31,561
Those poor engineers must be so frustrated. By hand tweaking AI content a bit, it's basically impossible to tell from actual human written-content. In fact, it IS human written content after all. AI's learn from human content, they don't create anything.
Hand tweaking AI content makes it non AI. Also Human written content has a reward associated with it when consumed unlike AI written content that is empty and not rewarding. It's not the bad english that is making AI content non human (it might even write perfect english content). It's the thought behind the content, that is inexistent with AI written articles and is a clear giveaway.
If you take AI content and make it Human by modifying it you're more or less just rewriting a piece of content, this could work.
If you take AI content and make it Human by modifying it you're more or less just rewriting a piece of content, this could work.
- Jan 22, 2011
- 3,691
- 781
Google openly denying AI content in words, but accepting so far, if that is user friendly.
If in real bot detects the manual written and machine written content, then there will be demand for writers!!
If in real bot detects the manual written and machine written content, then there will be demand for writers!!
- Oct 28, 2020
- 11,663
- 14,614
- Oct 9, 2013
- 3,471
- 14,475
This is not how machine learning works AT ALL. It's not a brute force technology in anyway way. It also doesn't update its self nonstop. Once the model is trained it doesn't have access to anymore data other than what you feed it through inference.
Machine learning is a prediction engine.
But. Let's ask chatgpt
<me>This is someone commenting on AI. Can you comment on what they said and correct any errors. "Exactly people forget or don't understand that AI (to simplify it) is a bruteforce technology. This type of tech works by taking information as input (google has all the information) and then using processing power to get to the desired result (google has all the processing power), also the AI updates itself nonstop(hard to be faster than that). Readable content that makes sense is not necessarily content that a Human would write, it will always be detectable by an AI that has google's infrastructure and power behind it, (it won't trick most humans either). "
<chatgpt> The comment makes several points about AI, but some of them are not entirely accurate. AI is not simply a "brute-force technology" that works by inputting information and using processing power to get a desired result. While some AI systems may use brute-force methods, many others use a variety of techniques, such as machine learning and neural networks, to learn from data and make predictions or decisions.
<me> So you're saying machine learning isn't a bruteforce technology?
<chatgpt> That is correct. While some AI systems may use brute-force methods, such as searching through large amounts of data to find a solution, many others use machine learning algorithms to learn from data and make predictions or decisions. Machine learning algorithms are a type of AI technology that involves training a model on a large dataset, allowing it to make predictions or decisions based on patterns and relationships it has learned from the data. This is different from brute-force methods, which simply involve trying many different possibilities until a solution is found.
I also asked "
What about his comment saying that the "AI updates its self nonstop". Is this true?
"
However, chatgpt seems to be having a bad day..
dami3n
Regular Member
- Nov 23, 2013
- 309
- 101
Let's say for the sake of the argument. They grab your "Title" and run it through multiple AI bots. Then compare that to your article. Yeah sounds expensive wasted server usage. It could find it out if you used google colab I guess. But yeah this sounds like a silly thing on googles part. If it's not spam and it gives great content, then that's all that should matter no?
patrickpogi123
Junior Member
- Mar 11, 2022
- 127
- 62
for people by the people... this phrase doesn't age well, overused just to appeal to public
Seems like chat gpt is kinda agreeing with me, some are based on bruteforce, the others have "bruteforce ideas/tech/concepts" in them (but then we will argue on the definition of bruteforce blahblahblah). There are some AI that can be in continous training in the sense that you can code auto retraining. All in all if you have a good model and it produces good quality content go for it, use whatever makes money, there is no need to be elitist when it comes to making money, my comments higher in this thread should be taken with a grain of salt (there might be some bias in them, I've mistaken human content for AI ).
- Oct 9, 2013
- 3,471
- 14,475
I had no idea you could do that yet.. Training is hard and needs to generally be done manually by setting epochs to make sure you're not overfitting the model.
Can you tell me more about how you code a model to auto train its self? That would be *incredible*. I'd like to know more about this.
hieu230592
Regular Member
- Jul 21, 2013
- 478
- 149
what if AI content is exactly what people search for lol
In the first place, Google need to be able to identify what is original content written by 'human' and AI. Currently, I don't there is an exact formula to identify AI usage in content generation. Especially if you use a human being to proof read before publishing.
Maybe the detection will be easier on image/ sound/ music generation whereby there is a fix pattern to capture. For content generation, this is going to take Google a lot of time before they can catch a pattern for content generation.
- Nov 25, 2017
- 11,570
- 7,545
Big LAUGH