Language Model Users Beware: 4 Pitfalls to Keep in Mind | Canada News Media
Connect with us

Science

Language Model Users Beware: 4 Pitfalls to Keep in Mind

Published

 on

Nowadays, language models like ChatGPT have been employed in a wide variety of tasks, ranging from fact-checking and email services to medical reporting and legal services.

Advertisements

While they are transforming our interaction with technology, it is important to remember that sometimes the information they give can be fabricated, conflicting, or old. As language models have this tendency to create false information, we need to be careful and aware of the problems that may arise when using them.

What Is a Language Model?

A language model is an AI program that can understand and create human language. The model is trained on text data to learn how words and phrases fit together to form meaningful sentences and convey information effectively.

Advertisements

The training is usually performed by enabling the model to predict the next word. After training, the model uses the learned ability to create text from a few initial words called prompts. For instance, if you provide ChatGPT with an incomplete sentence like “Techopedia is _____,” it will generate the following prediction: “Techopedia is an online technology resource that offers a wide range of articles, tutorials, and insights on various technology-related topics.”

The recent success of language models is primarily due to their extensive training in Internet data. However, while this training has improved their performance at many tasks, it has also created some issues.

Since the Internet contains incorrect, contradictory, and biased information, the models can sometimes give wrong, contradictory, or biased answers. It is, therefore, crucial to be cautious and not blindly trust everything generated by these models.

Advertisements

Hence, understanding the limitations of the models is vital to proceed with caution.

Hallucinations of Language Models

In AI, the term “hallucination” refers to the phenomenon where the model confidently makes incorrect predictions. It is similar to how people might see things that are not actually there. In language models, “hallucination” refers to when the models create and share incorrect information that appears to be true.

4 Forms of AI’s Hallucinations

Hallucination can occur in a variety of forms, including:

Fabrication: In this scenario, the model simply generates false information. For instance, if you ask it about historical events like World War II, it might give you answers with made-up details or events that never actually occurred. It could mention non-existent battles or individuals.

Factual inaccuracy: In this scenario, the model produces statements that are factually incorrect. For example, if you ask about a scientific concept like the Earth’s orbit around the Sun, the model might provide an answer that contradicts established scientific findings. Instead of stating the Earth orbits the Sun, the model might wrongly claim that the Earth orbits the Moon.

Sentence contradiction: This occurs when the language model generates a sentence that contradicts what it previously stated. For example, the language model might assert that “Language models are very accurate at describing historical events,” but later claim, “In reality, language models often generate hallucinations when describing historical events.” These contradictory statements indicate that the model has provided conflicting information.

Nonsensical content: Sometimes, the generated content includes things that make no sense or are unrelated. For example, it might say, “The largest planet in our solar system is Jupiter. Jupiter is also the name of a popular brand of peanut butter.” This type of information lacks logical coherence and can confuse readers, as it includes irrelevant details that are neither necessary nor accurate in the given context.

2 Key Reasons Behind AI’s Hallucinations

There could be several reasons that enable language models to hallucinate. Some of the main reasons are:

Data quality: Language models learn from a vast amount of data that can contain incorrect or conflicting information. When the data quality is low, it affects the model’s performance and causes it to generate incorrect responses. Since the models can not verify if the information is true, they may sometimes provide answers that are incorrect or unreliable.

Algorithmic limitations: Even if the underlying data is reliable, AI models can still generate inaccurate information due to inherent limitations in their functioning. As AI learns from extensive datasets, it acquires knowledge of various aspects crucial for generating text, including coherence, diversity, creativity, novelty, and accuracy. However, sometimes, certain factors, such as creativity and novelty, can take precedence, leading the AI to invent information that is not true.

Outdated Information

The language models like ChatGPT are trained on older datasets, which means they don’t have access to the latest information. As a result, the responses of these models may sometime be incorrect or outdated.

An example of how ChatGPT can present outdated information
When prompted with a question like “How many moons does Jupiter have?” NASA’s recent discovery indicates that Jupiter has between 80 and 95 moons. However, ChatGPT, relying on its data only up until 2021, predicts that Jupiter has 79 moons, failing to reflect this new finding.

This demonstrates how language models may provide inaccurate information due to outdated knowledge, making their responses less reliable. Additionally, language models can struggle to comprehend new ideas or events, further affecting their responses.

Therefore, when using language models for quick fact-checking or to get up-to-date information, it is essential to keep in mind that their responses may not reflect the most recent developments on the topic.

Impact of Context

Language models use previous prompts to enhance their understanding of user queries. This feature proves beneficial for tasks such as contextual learning and step-by-step problem-solving in mathematics.

However, it is essential to recognize that this reliance on context can occasionally lead to generating inappropriate responses when the query deviates from the previous conversation.

To get accurate answers, it is important to keep the conversation logical and connected.

Privacy and Data Security

Language models possess the capacity to utilize the information shared during interactions. Consequently, disclosing personal or sensitive information to these models carries inherent risks to privacy and security.

It is thus important to exercise caution and refrain from sharing confidential information when using these models.

The Bottom Line

Language models like ChatGPT have the potential to completely transform our interaction with technology. However, it is crucial to acknowledge the associated risks. These models are susceptible to generating false, conflicting, and outdated information.

They may experience “hallucinations” producing made-up details, factually incorrect statements, contradictory answers, or nonsensical responses. These issues can arise due to factors such as low data quality and inherent limitations of the algorithms employed.

The reliability of language models can be impacted by low data quality, algorithmic limitations, outdated information, and the influence of context.

Moreover, sharing personal information with these models can compromise privacy and data security, necessitating caution when interacting with them.

 

Source link

Continue Reading

News

Here’s how Helene and other storms dumped a whopping 40 trillion gallons of rain on the South

Published

 on

 

More than 40 trillion gallons of rain drenched the Southeast United States in the last week from Hurricane Helene and a run-of-the-mill rainstorm that sloshed in ahead of it — an unheard of amount of water that has stunned experts.

That’s enough to fill the Dallas Cowboys’ stadium 51,000 times, or Lake Tahoe just once. If it was concentrated just on the state of North Carolina that much water would be 3.5 feet deep (more than 1 meter). It’s enough to fill more than 60 million Olympic-size swimming pools.

“That’s an astronomical amount of precipitation,” said Ed Clark, head of the National Oceanic and Atmospheric Administration’s National Water Center in Tuscaloosa, Alabama. “I have not seen something in my 25 years of working at the weather service that is this geographically large of an extent and the sheer volume of water that fell from the sky.”

The flood damage from the rain is apocalyptic, meteorologists said. More than 100 people are dead, according to officials.

Private meteorologist Ryan Maue, a former NOAA chief scientist, calculated the amount of rain, using precipitation measurements made in 2.5-mile-by-2.5 mile grids as measured by satellites and ground observations. He came up with 40 trillion gallons through Sunday for the eastern United States, with 20 trillion gallons of that hitting just Georgia, Tennessee, the Carolinas and Florida from Hurricane Helene.

Clark did the calculations independently and said the 40 trillion gallon figure (151 trillion liters) is about right and, if anything, conservative. Maue said maybe 1 to 2 trillion more gallons of rain had fallen, much if it in Virginia, since his calculations.

Clark, who spends much of his work on issues of shrinking western water supplies, said to put the amount of rain in perspective, it’s more than twice the combined amount of water stored by two key Colorado River basin reservoirs: Lake Powell and Lake Mead.

Several meteorologists said this was a combination of two, maybe three storm systems. Before Helene struck, rain had fallen heavily for days because a low pressure system had “cut off” from the jet stream — which moves weather systems along west to east — and stalled over the Southeast. That funneled plenty of warm water from the Gulf of Mexico. And a storm that fell just short of named status parked along North Carolina’s Atlantic coast, dumping as much as 20 inches of rain, said North Carolina state climatologist Kathie Dello.

Then add Helene, one of the largest storms in the last couple decades and one that held plenty of rain because it was young and moved fast before it hit the Appalachians, said University of Albany hurricane expert Kristen Corbosiero.

“It was not just a perfect storm, but it was a combination of multiple storms that that led to the enormous amount of rain,” Maue said. “That collected at high elevation, we’re talking 3,000 to 6000 feet. And when you drop trillions of gallons on a mountain, that has to go down.”

The fact that these storms hit the mountains made everything worse, and not just because of runoff. The interaction between the mountains and the storm systems wrings more moisture out of the air, Clark, Maue and Corbosiero said.

North Carolina weather officials said their top measurement total was 31.33 inches in the tiny town of Busick. Mount Mitchell also got more than 2 feet of rainfall.

Before 2017’s Hurricane Harvey, “I said to our colleagues, you know, I never thought in my career that we would measure rainfall in feet,” Clark said. “And after Harvey, Florence, the more isolated events in eastern Kentucky, portions of South Dakota. We’re seeing events year in and year out where we are measuring rainfall in feet.”

Storms are getting wetter as the climate change s, said Corbosiero and Dello. A basic law of physics says the air holds nearly 4% more moisture for every degree Fahrenheit warmer (7% for every degree Celsius) and the world has warmed more than 2 degrees (1.2 degrees Celsius) since pre-industrial times.

Corbosiero said meteorologists are vigorously debating how much of Helene is due to worsening climate change and how much is random.

For Dello, the “fingerprints of climate change” were clear.

“We’ve seen tropical storm impacts in western North Carolina. But these storms are wetter and these storms are warmer. And there would have been a time when a tropical storm would have been heading toward North Carolina and would have caused some rain and some damage, but not apocalyptic destruction. ”

___

Follow AP’s climate coverage at https://apnews.com/hub/climate

___

Follow Seth Borenstein on Twitter at @borenbears

___

Associated Press climate and environmental coverage receives support from several private foundations. See more about AP’s climate initiative here. The AP is solely responsible for all content.

Source link

Continue Reading

Science

‘Big Sam’: Paleontologists unearth giant skull of Pachyrhinosaurus in Alberta

Published

 on

 

It’s a dinosaur that roamed Alberta’s badlands more than 70 million years ago, sporting a big, bumpy, bony head the size of a baby elephant.

On Wednesday, paleontologists near Grande Prairie pulled its 272-kilogram skull from the ground.

They call it “Big Sam.”

The adult Pachyrhinosaurus is the second plant-eating dinosaur to be unearthed from a dense bonebed belonging to a herd that died together on the edge of a valley that now sits 450 kilometres northwest of Edmonton.

It didn’t die alone.

“We have hundreds of juvenile bones in the bonebed, so we know that there are many babies and some adults among all of the big adults,” Emily Bamforth, a paleontologist with the nearby Philip J. Currie Dinosaur Museum, said in an interview on the way to the dig site.

She described the horned Pachyrhinosaurus as “the smaller, older cousin of the triceratops.”

“This species of dinosaur is endemic to the Grand Prairie area, so it’s found here and nowhere else in the world. They are … kind of about the size of an Indian elephant and a rhino,” she added.

The head alone, she said, is about the size of a baby elephant.

The discovery was a long time coming.

The bonebed was first discovered by a high school teacher out for a walk about 50 years ago. It took the teacher a decade to get anyone from southern Alberta to come to take a look.

“At the time, sort of in the ’70s and ’80s, paleontology in northern Alberta was virtually unknown,” said Bamforth.

When paleontogists eventually got to the site, Bamforth said, they learned “it’s actually one of the densest dinosaur bonebeds in North America.”

“It contains about 100 to 300 bones per square metre,” she said.

Paleontologists have been at the site sporadically ever since, combing through bones belonging to turtles, dinosaurs and lizards. Sixteen years ago, they discovered a large skull of an approximately 30-year-old Pachyrhinosaurus, which is now at the museum.

About a year ago, they found the second adult: Big Sam.

Bamforth said both dinosaurs are believed to have been the elders in the herd.

“Their distinguishing feature is that, instead of having a horn on their nose like a triceratops, they had this big, bony bump called a boss. And they have big, bony bumps over their eyes as well,” she said.

“It makes them look a little strange. It’s the one dinosaur that if you find it, it’s the only possible thing it can be.”

The genders of the two adults are unknown.

Bamforth said the extraction was difficult because Big Sam was intertwined in a cluster of about 300 other bones.

The skull was found upside down, “as if the animal was lying on its back,” but was well preserved, she said.

She said the excavation process involved putting plaster on the skull and wooden planks around if for stability. From there, it was lifted out — very carefully — with a crane, and was to be shipped on a trolley to the museum for study.

“I have extracted skulls in the past. This is probably the biggest one I’ve ever done though,” said Bamforth.

“It’s pretty exciting.”

This report by The Canadian Press was first published Sept. 25, 2024.

The Canadian Press. All rights reserved.

Source link

Continue Reading

News

The ancient jar smashed by a 4-year-old is back on display at an Israeli museum after repair

Published

 on

 

TEL AVIV, Israel (AP) — A rare Bronze-Era jar accidentally smashed by a 4-year-old visiting a museum was back on display Wednesday after restoration experts were able to carefully piece the artifact back together.

Last month, a family from northern Israel was visiting the museum when their youngest son tipped over the jar, which smashed into pieces.

Alex Geller, the boy’s father, said his son — the youngest of three — is exceptionally curious, and that the moment he heard the crash, “please let that not be my child” was the first thought that raced through his head.

The jar has been on display at the Hecht Museum in Haifa for 35 years. It was one of the only containers of its size and from that period still complete when it was discovered.

The Bronze Age jar is one of many artifacts exhibited out in the open, part of the Hecht Museum’s vision of letting visitors explore history without glass barriers, said Inbal Rivlin, the director of the museum, which is associated with Haifa University in northern Israel.

It was likely used to hold wine or oil, and dates back to between 2200 and 1500 B.C.

Rivlin and the museum decided to turn the moment, which captured international attention, into a teaching moment, inviting the Geller family back for a special visit and hands-on activity to illustrate the restoration process.

Rivlin added that the incident provided a welcome distraction from the ongoing war in Gaza. “Well, he’s just a kid. So I think that somehow it touches the heart of the people in Israel and around the world,“ said Rivlin.

Roee Shafir, a restoration expert at the museum, said the repairs would be fairly simple, as the pieces were from a single, complete jar. Archaeologists often face the more daunting task of sifting through piles of shards from multiple objects and trying to piece them together.

Experts used 3D technology, hi-resolution videos, and special glue to painstakingly reconstruct the large jar.

Less than two weeks after it broke, the jar went back on display at the museum. The gluing process left small hairline cracks, and a few pieces are missing, but the jar’s impressive size remains.

The only noticeable difference in the exhibit was a new sign reading “please don’t touch.”

The Canadian Press. All rights reserved.

Source link

Continue Reading

Trending

Exit mobile version