Tag Archives: tool

Nearly a third of employed Americans under 30 used ChatGPT for work: Poll – The Hill

Posted on by

More employed Americans have used the artificial intelligence (AI) tool ChatGPT for work since last year, with the biggest increase among the younger portion of the workforce, according to a Pew Research poll released Tuesday.  

The survey found that 31 percent of employed Americans between 18 and 29 surveyed in February said they have used ChatGPT for tasks at work, up from 12 percent who said the same last March.

The number of employed Americans who said they use ChatGPT for work decreased by age group. Twenty-one percent of employed adults aged 30 to 49 said they use it, up from 8 percent last year, and just 10 percent aged 50 and older said the same, up from only 4 percent last year.

Overall, the share of employed Americans who have used ChatGPT for work rose to double digits in the past year — reaching 20 percent based on the February survey, up from just 8 percent last March. But in general, most Americans still have not used ChatGPT, according to the survey.  

Twenty-three percent of Americans said they have used ChatGPT. That amount is on the rise from July, when 18 percent said the same.  

Use of ChatGPT has particularly spiked among younger adults. Forty-three percent of adults younger than 30 said they have used ChatGPT in the February survey, compared to 27 percent of adults 30 to 49, 17 percent of adults 50 to 64 and 6 percent of adults 65 and older.  

As the tool becomes more popular, OpenAI has also faced scrutiny about risks it presents about the spread of misinformation. OpenAI CEO Sam Altman faced questions about those risks and how it could impact the upcoming election when he testified before the Senate last year.  

Pew found that 38 percent of Americans said they do not trust the information from ChatGPT about the 2024 presidential election. Only 2 percent said they trust it a “great deal” or “quite a bit” and 10 percent said they have “some” trust in ChatGPT.  

The distrust of ChatGPT about information about the 2024 election was fairly evenly split between Republicans and Democrats.  

The survey also found that very few Americans, roughly 2 percent, said they have used the chatbot to find information about the presidential election.  

The survey is based on data from the American Trends Panel created by Pew Research Center and was conducted from Feb. 7-11. A total of 10,133 panelists responded out of 11,117 who were sampled. The margin of error for the full sample of 10,133 respondents is 1.5 percentage points.  

See original here:

Nearly a third of employed Americans under 30 used ChatGPT for work: Poll - The Hill

GPT-5 is ChatGPT’s next big upgrade, and it could be here very soon – Android Authority

Posted on by

Calvin Wankhede / Android Authority

TL;DR

OpenAIs ChatGPT has taken the world by storm, highlighting how AI can help with mundane tasks and, in turn, causing a mad rush among companies to incorporate AI into their products. GPT is the large language model that powers ChatGPT, with GPT-3 powering the ChatGPT that most of us know about. OpenAI has then upgraded ChatGPT with GPT-4, and it seems the company is on track to release GPT-5 too very soon.

According to a report from Business Insider, OpenAI is on track to release GPT-5 sometime in the middle of this year, likely during summer. Some enterprise customers are said to have received demos of the latest model and its related enhancements to ChatGPT, and they mention it to be really good, like materially better. These enterprise customers were showcased in a demo by OpenAI, which included use cases and data unique to the company.

Further, OpenAI is also said to have alluded to other as-yet-unreleased capabilities of the model, including the ability to call AI agents being developed by OpenAI to perform tasks autonomously.

The report clarifies that the company does not have a set release date for the new model and is still training GPT-5. Once training is complete, the model will be safety-tested internally. This includes red teaming the model, where it would be challenged in various ways to find issues before the tool is made available to the public. The safety testing has no specific timeframe for completion, so the process could potentially delay the release date.

The last major update to ChatGPT was a year ago with GPT-4. GPT-4 is faster and more accurate in its responses than GPT-3. The company also launched GPT-4 Turbo, which was made available to ChatGPT Plus subscribers. Before this report, GPT-5 was expected to take a while to train, develop, and test, potentially not releasing before 2025. The report gives us hope for an expedited release timeframe.

The report mentions that OpenAI hopes GPT-5 will be more reliable than previous models. Users have complained of GPT-4 degradation and worse outputs from ChatGPT, possibly due to degradation of training data that OpenAI may have used for updates and maintenance work.

In a recent interview with Lex Fridman, OpenAI CEO Sam Altman commented that GPT-4 kind of sucks when he was asked about the most impressive capabilities of GPT-4 and GPT-4 Turbo. He clarified that both are amazing, but people thought GPT-3 was also amazing, but now it is unimaginably horrible. Altman expects the delta between GPT-5 and 4 will be the same as between GPT-4 and 3. Altman commented, Maybe [GPT] 5 will be the pivotal moment, I dont know. Hard to say that looking forward. Were definitely looking forward to what OpenAI has in store for the future.

What are your expectations from GPT-5 and ChatGPT-5? What would you like to see improved? Let us know your thoughts in the comments below!

See more here:

GPT-5 is ChatGPT's next big upgrade, and it could be here very soon - Android Authority

Meta’s New Image-Generating AI Is Trained on Your Instagram and Facebook Posts

Posted on by

Earlier this week, Meta announced a new AI image generator dubbed

Cashing In

Earlier this week, Meta announced a new AI image generator dubbed "Imagine with Meta AI."

And while it may seem like an otherwise conventional tool meant to compete with the likes of Google's DALL-E 3, Diffusion, and Midjourney, Meta's underlying "Emu" image-synthesis model has a dirty little secret.

What's that? Well, as Ars Technica points out, the social media company trained it using a whopping 1.1 billion Instagram and Facebook photos, per the company's official documentation — the latest example of Meta squeezing every last drop out of its user base and its ever-valuable data.

In many ways, it's a data privacy nightmare waiting to unfold. While Meta claims to only have used photos that were set to "public," it's likely only a matter of time until somebody finds a way to abuse the system. After all, Meta's track record is abysmal when it comes to ensuring its users' privacy, to say the least.

So Creative

Meta is selling its latest tool, which was made available exclusively in the US this week, as a "fun and creative" way to generate "content in chats."

"This standalone experience for creative hobbyists lets you create images with technology from Emu, our image foundation model," the company's announcement reads. "While our messaging experience is designed for more playful, back-and-forth interactions, you can now create free images on the web, too."

Meta's Emu model uses a process called "quality-tuning" to compare the "aesthetic alignment" of comparable images, setting it apart from the competition, as Ars notes.

Other than that, the tool is exactly what you'd expect. With a simple prompt, it can spit out four photorealistic images of skateboarding teddy bears or an elephant walking out of a fridge, which can then be shared on Instagram or Facebook — where, perhaps, they'll be scraped by the next AI.

Earlier this year, Meta's president for global affairs Nick Clegg told Reuters that the company has been crawling through Facebook and Instagram posts to train its Meta AI virtual assistant as well as its Emu image model.

At the time, Clegg claimed that Meta was excluding private messages and posts, avoiding public datasets with a "heavy preponderance of personal information."

Instead of immediately triggering a massive outcry and lawsuits over possible copyright infringement like Meta's competitors, the social media company can crawl its own datasets, which come courtesy of its users and its expansive terms of service.

But relying on Instagram selfies and Facebook family albums comes with its own inherent risks, which may well come back to haunt the Mark Zuckerberg-led social giant.

More on Meta: Facebook Has a Gigantic Pedophilia Problem

The post Meta's New Image-Generating AI Is Trained on Your Instagram and Facebook Posts appeared first on Futurism.

More:
Meta's New Image-Generating AI Is Trained on Your Instagram and Facebook Posts