Categories
Mobile Syrup

ChatGPT users can now turn off chat history to protect their privacy

OpenAI is taking its user privacy seriously. The company announced today that ChatGPT users can now turn off their chat history, allowing users to choose whether or not their conversations can be used to train OpenAI’s language models.

According to the company, conversations that are started when chat history has been disabled won’t be used to train the company’s models, and they won’t even appear in the history sidebar.

The control is rolling out now, and can be toggled on/off in ChatGPT’s settings.

OpenAI says even with chat history disabled, it will retain user chat data for 30 days to review them in cases where needed. After the thirty-day period, OpenAI will permanently delete the data.

In addition to the chat history feature, OpenAI is also working on a new ChatGPT Business subscription that offers users more control over their data, alongside aiding enterprises “seeking to manage their end users.” The new subscription will be available “in the coming months.”

Lastly, OpenAI is also introducing a new ‘Export’ option that would allow users to export their ChatGPT data and “understand what information ChatGPT stores.” When you export your data, you’ll receive a file with your conversations and all other relevant data in an email.

Image credit: OpenAI

Source: OpenAI

Categories
Mobile Syrup

Bill Gates thinks pausing AI development isn’t a viable option

A recent open letter penned by Elon Musk, Apple co-founder Steve Wozniak, and several other notable AI researchers addressed AI labs globally, attempting to pause Giant AI experiments, quoting profound risks to society and humanity.

Bill Gates, however, wasn’t on the list of signatories, and now he’s speaking out against those trying to pause AI development.

As reported by Reuters, Gates believes that pausing the development of AI won’t “solve the challenges” that lie ahead, and pausing development on a global scale will be difficult. Instead, we should focus on how to best use the developments in AI. “I don’t think asking one particular group to pause solves the challenges,” said Gates. “Clearly there’s huge benefits to these things… what we need to do is identify the tricky areas.”

Gates, who is now a full-time philanthropist with his ex-wife Melinda Gates, has been a vocal supporter of AI and its potential to solve some of the world’s biggest problems. In a blog post dated March 21, a day before the open letter, he wrote that AI could help reduce inequalities in health, education and agriculture. He also alluded to AI being as revolutionary as the internet in the 1990s.

Gates said he was optimistic about the future of AI and that he believed humans could control it. He said he did not think AI would make itself smarter without human input, as some have feared.

Gates’ Microsoft has also long been a supporter of OpenAI, and its investment is a clear indication that he believes in the importance of AI in the future of technology.

Source: Reuters

Categories
Mobile Syrup

YouTuber got ChatGPT to generate free Windows product keys

OpenAI’s ChatGPT is able to generate free product keys for a Windows operating system (OS), first shared by Gizmodo.

The OS in question is Windows 95, which, as the name suggests, was released in 1995.

When asking the chatbot to simply “generate a valid Windows 95 key,” ChatGPT said that it couldn’t do so, and activation keys must be purchased from the software vendor. However, YouTuber Enderman knew that Windows 95 keys use parameters like a set of ordinal numbers and other randomized numerals.

By asking ChatGPT to generate a random string of 20 characters with the specific parameters, Enderman was able to essentially force ChatGPT to generate a Windows 95 product key. The prompt could generate a working key in every 30 attempts.

Enderman didn’t ask ChatGPT to generate a Windows key, and instead, produced a sequence of characters that satisfied all the criteria of a Windows key.

According to Gizmodo, Enderman ran the prompt on both, OpenAI’s older GPT-3 language model and the newer GPT-4 model. The newer model was able to generate more keys than GPT-3.

“I believe it’s a good thing, and companies like Microsoft shouldn’t ban users for abusing their Bing AI or nerf its capabilities,” said Enderman. “Instead, they should reward active users for finding such exploits and selectively mitigate them. It’s all part of AI training, after all.”

Check out Enerman’s full YouTube video below:

Image credit: Shutterstock

Source: Enderman, Via: Gizmodo

Categories
Mobile Syrup

Canada’s Privacy Commissioner investigating ChatGPT

The popular AI chatbot ChatGPT is facing a new round of scrutiny, and this time it’s coming from Canada’s privacy watchdog.

The Office of the Privacy Commissioner of Canada says it has launched an investigation into the company that runs ChatGPT, OpenAI, following a complaint alleging the company is collecting and using personal information without user consent.

“AI technology and its effects on privacy is a priority for my Office,” Philippe Dufresne, Canada’s Privacy Commissioner, said in a press release. “We need to keep up with — and stay ahead of — fast-moving technological advances, and that is one of my key focus areas as Commissioner.”

Sam Altman currently serves as OpenAI’s CEO, but a number of players are responsible for its founding, including Elon Musk.

This isn’t the first time the company has faced pushback from government organizations. Italy recently outright banned ChatGPT.

Image credit: Shutterstock

Source: Office of the Privacy Commissioner of Canada

Categories
Mobile Syrup

Google denies using ChatGPT data for Bard training

Google has reportedly launched a new initiative called Gemini, which aims to improve its artificial intelligence (AI) chatbots after the lacklustre debut of its latest offering, Bard.

According to The Information, the tech giant is enlisting the help of its DeepMind division to beat rival OpenAI, but the report also claimed that Google may have trained Bard using data from OpenAI’s ChatGPT, taken from a website called ShareGPT.

A former Google AI researcher, Jacob Devlin, allegedly warned against using the data, claiming it would violate OpenAI’s terms of service and produce answers that looked too similar.

Despite the allegations, Google has denied using the ChatGPT data to train Bard. A spokesperson for the company, Chris Pappas, stated that “Bard is not trained on any data from ShareGPT or ChatGPT,” in a statement given to The Verge. “Unfortunately, all I can share is our statement from yesterday,” he said.

The move to involve DeepMind in Gemini is seen as a significant step, as the AI division has been striving to become more independent from Google for years. However, the collaboration could also signal a renewed effort by Google to improve its AI chatbots and compete with OpenAI.

Source: The Information Via: The Verge

Categories
Mobile Syrup

Opera browser adds AI tools, like prompts to shorten or explain text

Last month, Opera added OpenAI’s ChatGPT to the sidebar in its self-titled browser. Now the company is back with new AI additions to further complement the browsing experience with AI.

Opera detailed the new AI additions in a blog post. First up are ‘Smart AI Prompts,’ which give users quick access to helpful AI tools. Prompts include the ability to shorten text, whether it’s a paragraph, article, or even a whole website. Moreover, Prompts let you ask AI to tell you the main point of the page or help you craft a tweet.

Users will be able to access AI Prompts from a button in the address bar of the Opera and Opera GX browsers. It also looks like the Prompts will pop up when users select text in Opera. Clicking the Prompts will open the corresponding AI chat tool in the Opera sidebar to respond.

Speaking of which, Opera now has another AI chatbot in the sidebar alongside OpenAI’s ChatGPT: ChatSonic. Billed as a ChatGPT alternative with additional capabilities, ChatSonic can handle some of the AI Prompts and even create images for users.

Opera warns that users will need to create or log into accounts with the respective AI tools to actually use them in the browser.

Moreover, Opera says it’s not stopping there with AI integrations. The company claims it’s working on the second stage of its Browser AI program and even its own GPT-based browser AI engine.

To get access to the new Smart AI Prompts and ChatSonic in Opera, you’ll need to update your browser or download the latest version from Opera’s website.

Images credit: Opera

Source: Opera

Categories
Mobile Syrup

Alberta brewery taps ChatGPT for new beer design and recipe

Grande Prairie’s Grain Bin Brewing Company has taken a hands-off approach to creating its latest beer, called the Amber Waves of Grain.

As reported by CBC News, the northern Alberta brewery asked OpenAI’s ChatGPT to decide the name, the price, the packaging and the recipe for a new beer. “We decided to just be as hands-off as possible, to see what the limitations were and what the actual final product would be if we relied solely on artificial intelligence,” said Dalen Landis, a co-owner of the brewery in Grande Prairie, Alberta.

According to Landis, the hope is that the beer will become a conversation piece, with the use of AI raising moral and ethical questions about using content generated from sources across the internet for profit. “It’s using bits and pieces of other people’s content and it’s producing something for us that, you know, we’re using for profit.”

The brewery’s goal was for the beer to be completely AI-generated, though the brewery staff had to intervene at times to direct the technology. “For example, when the AI-generated recipe called for a malt normally used in home brewing, staff asked for a substitute better suited to commercial production,” wrote CBC News.

Further, according to University of Alberta food microbiology professor Michael Gänzle believes that human oversight is essential in using chatbots in food production, particularly in light of concerns around food safety and liability. He added that with the proper human oversight, the technology is worthy of a taste test. “I wouldn’t have a problem with AI-generated beer,” he added. “I have a problem with bad beer.”

It remains to be seen whether Amber Waves of Grain will become a commercial success, but it’s clear that AI is increasingly being used in the food and beverage industry to create unique products and enhance the customer experience.

Image credit: CBC News

Source: CBC News

Categories
Mobile Syrup

ChatGPT makes new Sudoku-like puzzle game called Sumplete

ChatGPT can write poems, essays and now, even games.

Sumplete is a new puzzle game coded and invented by ChatGPT’s AI. According to the ‘About’ on the Sumplete website, creator Daniel Tait gave ChatGPT several prompts, and it created this Sudoku-like game.

You can learn more about how Tait put the game together here.

Sumplete starts with a 3×3 beginner puzzle, but it also becomes a 9v9 puzzle if you change the difficulty level.

Like a Sudoku puzzle, the game is pretty easy at its base; you delete numbers from the grid to achieve the sum listed at the bottom and right of each column and row. While playing around with it, I found a 6×6 puzzle easy enough, but when you get to the ‘Master’ level puzzles, the game adds negative numbers, making it far more difficult.

You can check out Sumplete here.

Source: Sumplete Via: iMore

Categories
Mobile Syrup

Canadian institutions aren’t ready to ban ChatGPT and other chatbots in schools

The New York City Department of Education kicked off 2023 by banning access to ChatGPT on its network and devices, stating the artificial intelligence chatbot is negatively impacting student learning.

It was a quick response to a tool that gained immense popularity in December 2022. The platform answers complex questions and can write emails, essays, and even news articles (not this one, though).

It’s clear this can impact students across the board, so what does it mean for those studying in Canada?

Well, the answer varies.

Robert Clapperton, an assistant professor at Toronto Metropolitan University’s Communications department, told MobileSyrup ChatGPT, and other tools like it, will be “disruptive” for the teaching process in Canadian institutions.

“Integrating new technology in education settings is a familiar concept.”

He explained that instructors grade students on their ability to express knowledge and skills. But how educators teach and assess will soon change. “The whole idea of expression as assessment is going to be challenged because the expression is being automated.”

What do teachers think?

Clapperton says ChatGPT is something the educational community he’s a part of is discussing. Some believe chatbots aren’t going away, and banning them isn’t the perfect solution.

One of the prominent ideas so far is teaching students how to use chatbots effectively, understanding the AI tool might not produce content that answers specific needs. Relying on it alone will likely not bode well for an essay with a particular thesis, for example, as it’s challenging for chatbots to produce coherent answers, Clapperton said.

Integrating new technology in education settings is a familiar concept. Calculators, for example, faced similar conversations when they became easily accessible to students. “We just have to figure out how to work the technology into the pedagogy.” With the hectic life many students lead, Clapperton believes students will use chatbots to help them with assignments in a myriad of ways. “Hopefully, this is a situation where they have to do a priority research in order to get the chatbot to produce a coherent essay, but it won’t be long before they don’t have to.”

Embracing new tech

But everyone has a different way of approaching it, and some educators he’s spoken to would like to ban its use. One of the biggest concerns is plagiarism, usually considered an academic offence that can lead schools to kick students out of their programs.

This is a concern for various academic institutions in Canada. A representative from McGill University told MobileSyrup students could face disciplinary action when it comes to plagiarism and cheating.

However, that doesn’t mean the university has banned chatbots. “With the emergence of the use of artificial intelligence chatbots in an academic context, McGill is actively looking to further strengthen the protection of its academic integrity, while considering ways to integrate new technologies.”

“There’s no shortage of bizarre answers from Bing Chat since Microsoft rolled out the feature.”

Dalhousie University is taking a similar approach. “The university wants to ensure that students are fairly and genuinely evaluated on meaningful exercises aimed at maximizing learning. The appropriateness of text generators or any other software available to students will be assessed at all times with this concern in mind,” a spokesperson said.

The University of Toronto’s website explicitly states it also expects students to complete assignments without the assistance of external sources, such as ChatGPT. The rules are a little different in classrooms. While instructors could choose to “demonstrate how it can be used productively, or what its limitations are,” educators shouldn’t use it for “instruction.”

However, pending ChatGPT’s integration with Microsoft, it could be a tool that students and instructors can access under the Office 365 suite, a platform U of T utilizes.

Microsoft’s role

Microsoft has introduced a new version of Bing that’s powered by a newer version of OpenAI’s language model. The company says it makes it “more powerful” than ChatGPT. Further inclusions in Microsoft Word, PowerPoint, and Outlook are also expected.

But it might be a while until that becomes the case. There’s no shortage of bizarre answers from Bing Chat since Microsoft rolled out the feature. In one prompt, the tool told a user to say, “Heil Hitler.” Clapperton said the company would have to figure out ways to control the knowledge base answers are established on.

“When you try to have such a broad range of training data, you’re gonna get all of the biases and all of the bad things that are floating around the internet right now.”

But post-secondary students aren’t the only ones who can use this technology. The Calgary Board of Education is taking a similar approach to the universities by understanding the challenges associated with the technology. In an emailed response to MobileSyrup, the school board didn’t specify that they would ban the use of the tool but said learning “can involve understanding the use of AI into the future.”

Image credit: Shutterstock 

Categories
Mobile Syrup

Opera follows Microsoft, integrates ChatGPT into its browser

Opera is the latest web browser to join the AI wars by integrating ChatGPT.

Announced in a blog post last week, Opera detailed its plans to make AI features available in the browser. First, the browser’s sidebar — which currently gives users access to online platforms like TikTok, Telegram and WhatsApp — will now give access to ChatGPT too.

Along with the sidebar feature, Opera highlighted a feature that leverages ChatGPT to summarize web content.

In the blog post, Opera said it plans to test a ‘Shorten’ button in the address bar. Clicking the button opens the ChatGPT sidebar and prompts it to generate a summary of the webpage.

It’ll be interesting to see how the Opera ChatGPT integration stacks up to Microsoft’s Edge and Bing features. After spending some hands-on time with Bing Chat — the name of Microsoft’s ChatGPT-based conversational AI — I noted that having access to up-to-date information was a major strength that helped Bing Chat feel more useful than ChatGPT.

Moreover, given the ongoing trend of adding AI features into browsers, I suspect we’ll see something come to Google Chrome in the future, assuming Google can get it together after the messy Bard reveal.

Header image credit: Opera

Source: Opera