Skip to main content

The model collapse can danger LLM.


An example of model collapse in an image-generating AI. (FreeThink,“Model collapse” threatens to kill progress on generative AIs)

"When AI eats its own product, it gets sick." (FreeThink)

The images introduce how the AI. That surrounds the same data and causes corruption in the data's entirety. When data travels in wires, there is always some kind of disturbance. Those electromagnetic and physical anomalies cause a lost bit or two in the data that travels in the wire. The loss of one or two bits doesn't mean much. When the data surrounds the system thousands of times. That causes a very bad change in the data flow. And sooner or later, the data mass is corrupted and turned into another one. 

The corruption in data is the same thing as the model collapse in the data flow. 

When we use large language models LLM's we use computer programs that can search and connect data by using certain algorithms. That algorithm searches for details from search and then connects data from sources that fill certain rules. That makes it possible for the LLM can make almost suitable academic texts and translate homepages. The problem with the LLM is that it doesn't think. 

The LLM sees letters as ASCII codes and searches for similarities in the ASCII codes that it sees in homepages. So the LLM doesn't know the meaning of the words that it uses to generate new texts. The LLM collects words that look like the same ones that are in the search field. And then the AI connects those words into the new entirety. 

In that model, the LLM starts to surround the data. That means the LLM recycles texts that it made itself. And that thing forms the corruption in the code, that the LLM recycles. When LLM starts to use its texts and images as sources, that can cause problems with data, that the LLM uses. 

One of the biggest problems is that the AI can start to recycle its texts. Or the AI can interconnect the data, that other AI's generate. And when the AI generates data, it just connects sources to the new entireties. 

The AI doesn't know what words mean. And that makes it easy to manipulate its answers changing the involvements of those homepages. 

The problem is this. When the AI selects sources, it uses the homepages from the search engine. The search engine doesn't make a difference does the click come from the LLM or the human? And the LLM can raise some homepages pagerank. It's possible. Some homepages involve cool-looking information but having some kind of scientific-looking heading or title in the search path can raise the homepage that has nothing to do with the thing, that maybe the student searches. 

And that makes the AI give text, there is nothing to do with studies. In the worst case, the person who does not know the searched topics puts that AI-created crab on Wikipedia or something like that. In the worst case, the AI starts to recycle that data. And in some other cases, the AI starts to connect data, that other AI's created. In those cases the data that AI connects and rewrites starts to corrupt. 

https://www.freethink.com/robots-ai/model-collapse-synthetic-data

Comments

Post a Comment

Popular posts from this blog

The hydrogen-burning supernovas are interesting models.

"Researchers discovered a significant magnesium anomaly in a meteorite’s dust particle, challenging current astrophysical models and suggesting new insights into hydrogen-burning supernovas. (Artist’s concept.)Credit: SciTechDaily.com" (ScitechDaily, Rare Dust Particle From Ancient Extraterrestrial Meteorite Challenges Astrophysical Models) If the star is too heavy when its fusion reaction starts, it can detonate just at that moment, when its fusion starts. If the collapsing nebula is heavy enough, it can form a black hole straight from the nebula. But if the nebula's gravity is too heavy to  form  the blue giant or too  small  it can collapse  straight  into a black hole . If  the forming star is a little bit larger than the blue supergiants. It can explode immediately when the fusion starts.    The theory of hydrogen-burning supernovas consists  model  of the giant stars that explode immediately after their fusion starts. When the...
Post a Comment
Read more

The ancient galaxy mirrors the Milky Way.

"This image shows the galaxy REBELS-25 as seen by the Atacama Large Millimeter/submillimeter Array (ALMA), overlaid on an infrared image of other stars and galaxies. The infrared image was taken by ESO’s Visible and Infrared Survey Telescope for Astronomy (VISTA). In a recent study, researchers found evidence that REBELS-25 is a strongly rotating disc galaxy existing only 700 million years after the Big Bang. This makes it the most distant and earliest known Milky Way-like galaxy found to date. Credit: ALMA (ESO/NAOJ/NRAO)/L. Rowland et al./ESO/J. Dunlop et al. Ack.: CASU, CALET" (ScitechDaily, Astronomers Baffled by Ancient Galaxy That Mirrors Modern Milky Way) Researchers found the youngest Milky Way-type galaxy. The distance to the galaxy is enormous. And the light that comes from that galaxy named REBELS-25 comes from the Universe that is only 700 million years old. The distance to that galaxy is enormous about 236 billion light years. And that means it's a very dista...
Post a Comment
Read more

Transcendence, or the ability to transcendent thinking may grow in teen's brains.

   "New research has discovered that transcendent thinking, which involves analyzing the broader implications of situations, can foster brain growth in adolescents. This form of thinking enhances brain network coordination, impacting developmental milestones and future life satisfaction. The study emphasizes the need for education that encourages deep, reflective thought, underscoring the critical role of adolescents in their own brain development". (ScitechDaily, Scientists Discover That “Transcendent” Thinking May Grow Teens’ Brains) "Scientists at  USC Rossier School of Education’s Center for Affective Neuroscience, Development, Learning and Education (CANDLE) have discovered that adolescents who grapple with the bigger meaning of social situations experience greater brain growth, which predicts stronger identity development and life satisfaction years later". (ScitechDaily, Scientists Discover That “Transcendent” Thinking May Grow Teens’ Brains) The transcendenc...
Post a Comment
Read more