OpenAI CEO Sam Altman and several other company executives hosted an ask-me-anything (AMA) session on Thursday. The session was hosted on the social networking platform Reddit and users were told to ask questions about the AI firm’s products such as ChatGPT or general queries about artificial intelligence (AI) and artificial general intelligence (AGI). During the session, Altman said that GPT-5 will not be released this year, however, the company plans to introduce “some very good releases” before the end of 2024.

OpenAI Staff Host AMA on Reddit

The AMA session was hosted on the ChatGPT subreddit. Calling it “our Reddit launch”, several executives including Altman, OpenAI CPO Kevin Weil, SVP of Research Mark Chen, VP Engineering ​​Srinivas Narayanan, and Chief Scientist Jakub Pachocki participated in the question-and-answer post. The official X (formerly known as Twitter) handle of OpenAI also posted about the Reddit AMA.

Answering a question about the timeline for GPT-5 or its equivalent’s release, Altman said, “We have some very good releases coming later this year! nothing that we are going to call gpt-5, though.” This seems on par with what multiple reports have confirmed with most expecting OpenAI to release the next flagship model sometime in 2025.

Another user asked about the value that SearchGPT or the ChatGPT Search feature brings, Altman said that he finds it to be a faster and easier way to get to the information. He also highlighted that the web search functionality will be more useful for complex research. “I also look forward to a future where a search query can dynamically render a custom web page in response,” he added.

Weil also fielded several questions from users. One asked about the delay in Sora, to which the OpenAI CPO said that the delay was caused due to additional time taken to perfect the model, getting safety and impersonation right, and the need to scale compute. However, he did not reveal a date for its launch.

Weil also highlighted that the ‘o’ series AI models, such as GPT-4o and o1-preview, will become a mainstay in the company’s lineup and will make an appearance even after the release of GPT-5. Additionally, he also revealed that the ChatGPT Advanced Voice Mode could be tweaked to add a singing voice to the AI.

One user suggested, “Can we please get a hands-free way to end a voice conversation with ChatGPT? Using the action button shortcut on iPhone, but the voice chat needs manually dismissed every time.” Weil seemed to be impressed with the idea and said, “I love this idea. Sharing with the team now!”

OpenAI SVP or Research Mike Chen also answered an important user question about AI hallucination. Explaining why hallucinations from AI models are not completely gone, he called it a fundamentally hard problem. This is because the AI models learn from human-written text, and humans can often make errors, which then are added to the core dataset of the large language models (LLMs).

“Our models are improving at citing, which grounds their answers in trusted sources, and we also believe that RL will help with hallucinations as well – when we can programmatically check whether models hallucinate, we can reward it for not doing so,” Chen added.

Microsoft on Monday said that its vice president of GenAI research, Sebastien Bubeck, is leaving the company to join ChatGPT-maker OpenAI.

It was unclear what role Bubeck would assume at the Microsoft-backed AI startup.

“Sebastian has decided to leave Microsoft to further his work toward developing AGI,” a spokesperson for Microsoft said, adding that the company looks forward to continuing their relationship through Bubeck’s work with OpenAI.

Bubeck did not immediately respond to a Reuters request for confirmation.

Most of Bubeck’s coauthors who worked on a research paper on Microsoft’s Phi LLMs, which are smaller than traditional large language models (LLMs), remain at Microsoft and plan to continue developing the models, according to the Information, which first reported his departure.

The development follows a series of departures from OpenAI, including that of longtime chief technology officer Mira Murati in September.

CEO Sam Altman has denied any link between the departures and a planned restructuring of the company.

© Thomson Reuters 2024

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)

OpenAI said a group with apparent ties to China tried to carry out a phishing attack on its employees, reigniting concerns that bad actors in Beijing want to steal sensitive information from top US artificial intelligence companies.

The AI startup said Wednesday that a suspected China-based group called SweetSpecter posed as a user of OpenAI’s chatbot ChatGPT earlier this year and sent customer support emails to staff. The emails included malware attachments that, if opened, would have allowed SweetSpecter to take screenshots and exfiltrate data, OpenAI said, but the attempt was unsuccessful.

“OpenAI’s security team contacted employees who were believed to have been targeted in this spear phishing campaign and found that existing security controls prevented the emails from ever reaching their corporate emails,” OpenAI said.

The disclosure highlights the potential cybersecurity risks for leading AI companies as the US and China are locked in a high-stakes battle for Artificial Intelligence supremacy. In March, for example, a former Google engineer was charged with stealing AI trade secrets for a Chinese firm.

China’s government has repeatedly denied allegations by the US that organisations within the country perpetrate cyberattacks, accusing external parties of organising smear campaigns.

OpenAI revealed the attempted phishing attack as part of its latest threat intelligence report, outlining its efforts to combat influence operations around the world. In the report, OpenAI said it took down accounts from groups with links to Iran and China that used AI for coding assistance, conducting research and other tasks.

© 2024 Bloomberg LP

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)

Japanese telecom company SoftBank’s Vision Fund will invest $500 million (roughly Rs. 4,190 crore) in OpenAI’s latest funding round, The Information reported on Monday, citing a person familiar with the deal.

SoftBank declined to comment, while OpenAI did not immediately respond.

The company at the heart of the Artificial Intelligence (AI) boom is raising $6.5 billion (roughly Rs. 54,478 crore) in the form of convertible notes, Reuters exclusively reported earlier this month.

Apple reportedly dropped out of plans to participate in the large funding round, which currently values the AI startup at $150 billion (roughly Rs. 12,57,187 crore) before the SoftBank investment.

However, the valuation will be contingent on whether the ChatGPT-maker can upend its corporate structure and remove a profit cap for investors.

The deal represents SoftBank’s first investment in the Sam Altman-led firm, the Information report said.

Reuters reported on Wednesday that OpenAI is working on a plan to restructure its core business into a for-profit corporation that will no longer be controlled by its non-profit board.

© Thomson Reuters 2024

Sam Altman denied on Thursday that there was any link between the departure of three senior OpenAI executives and a planned restructuring of the company which he said the board had been considering for several months.

The world-leading AI firm’s longtime chief technology officer Mira Murati abruptly announced her departure on Wednesday. Within hours, two senior research executives Barret Zoph and Bob McGrew revealed they were also leaving the company.

On the same day, Reuters reported OpenAI was working on a plan to restructure its core business into a for-profit benefit corporation that will no longer be controlled by its non-profit board, in a move aimed at making the company more attractive to investors.

Speaking on stage at the Italian Tech Week conference in Turin, OpenAI CEO Altman said that “some stuff” reported around the executives’ departures was inaccurate, adding that the personnel changes were unrelated to the restructuring.

“That’s totally not true,” he told event organiser John Elkann, the chairman of car companies Ferrari and Stellantis.

“A lot of the stuff I saw was also just totally wrong, but we have been thinking about that (restructuring), our board has, for almost a year, independently, as we think about what it takes to get to our next stage.”

Altman praised the departing executives, adding he was excited to simplify the company’s structure and work more closely with technical staff.

“I have not been as involved in the tech recently as other things, because there’s been so much going on, I’m excited to do that,” he said.

“This will be, hopefully, a great transition for everyone involved, and OpenAI will be stronger for it, as we are for all our transitions.”

Details of the proposed restructure highlight significant changes behind the scenes at one of the world’s top AI companies.

Sources told Reuters the plans were still being negotiated between lawyers and shareholders at the time of writing, with the timeline for completion still uncertain.

© Thomson Reuters 2024

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)

Three top technical leaders at Microsoft-backed OpenAI quit on Wednesday, the latest in a string of executive departures this year coming at a time of flux for the ChatGPT maker.

Chief Technology Officer Mira Murati, VP Research Barret Zoph and Chief Research Officer Bob McGrew all announced their departures via X on Wednesday afternoon.

The San Francisco-based Artificial Intelligence (AI) startup is negotiating a new $6.5 billion (roughly Rs. 54,398 crore) financing round valuing the company at $150 billion (roughly Rs.12,55,360 crore), contingent on the company upending its corporate structure.

The company plans to restructure to a for-profit benefit corporation and will give CEO Sam Altman an equity stake, Reuters reported on Wednesday. Currently, a non-profit board controls the for-profit entity, an unusual structure which led to members of the non-profit board ousting Altman in November 2023 over a breakdown in communication and loss of trust. He was reinstated after five days.

The funding round has not closed yet and the company is in the process of finalising it.

It was unclear whether the executives’ departure could affect the ongoing fundraise. Some fundraising documents contain a “material adverse change” clause which allows investors to withdraw from a deal if the company encounters anything that could have a significant negative impact.

Murati is still working at OpenAI while she negotiates her exit from the company, according to a source close to Murati. She has been part of the ChatGPT maker for 6-1/2 years, and briefly served as CEO in November when the board temporarily Altman.

According to her LinkedIn profile, Murati joined OpenAI as the “VP of Applied AI and Partnerships” in December 2020 and was promoted to CTO in May 2022. Prior to OpenAI, she worked at virtual and augmented reality startup Leap Motion and at Tesla.

As CTO, Murati frequently appeared alongside Altman as the public face of the ChatGPT maker. When OpenAI launched its GPT-4o model in May, capable of having realistic voice conversation, Murati led the presentation.

“I’m stepping away because I want to create the time and space to do my own exploration,” Murati wrote a post on X.

Murati, Zoph and McGrew are the latest in a string of OpenAI executive departures. In August, OpenAI co-founder John Schulman said on X that he had joined rival AI company Anthropic and another co-founder, Greg Brockman, also said on X he was taking a sabbatical through the end of the year. A third co-founder, chief scientist Ilya Sutskever, left OpenAI in May.

On X, Altman said Murati did not inform him in advance of her departure, explaining to him that she wanted to do it when the company was in an upswing and that “there is never a good time”. In the same post, Altman also announced a slew of internal promotions: Matt Knight to chief information security officer, Josh Achiam to head of mission alignment and Mark Chen to senior vice president of Research.

© Thomson Reuters 2024

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)

Big tech is going big on generative AI. Almost every tech product is set to add generative AI features in some form or the other. Amazon Alexa, the popular virtual assistant, is also getting its dose of conversational AI tools that can help accelerate the company’s vision to create a great AI-powered personal assistant. “It’s now a question of just how do we implement it,” said Dave Limp, senior VP of devices and services, in a recent media briefing at the company’s headquarters in Seattle, US.

The goal is to enable Alexa to have more natural conversations with users. Generative AI can help achieve that goal even faster. During the briefing, Limp also explained why Amazon cannot simply take ChatGPT and throw it into Alexa. He explained that such large language models tend to “hallucinate” from time to time. For a virtual assistant like Alexa, that could spell trouble. Customers rely on Alexa’s capabilities to offer accurate information — be it prices, weather, or simply turning on your smart home equipment. Limp also confirmed that a lot of generative AI models in the background are already in production.

Limp revealed how consumers are now using Alexa more than ever. He said that year on year, Alexa engagement is up over 30 percent. “It has become their new normal, and it has become a new habit,” he said. Consumers are listening to music and podcasts, shopping, building lists, setting up alarms, etc. Amazon is also working on a bunch of new Alexa-powered products that are “basically in the oven and baking, and ready to come put.”

We spoke to Michele Butti, Head of Alexa International, to understand how Alexa is doing in India, when Alexa will get ChatGPT-like features, and what the company is planning for its Indian audience in the near future.

Note: Some responses have been condensed and slightly edited for clarity.

Gadgets 360: Can you talk about Alexa’s engagement in countries like India?

Michele Butti: Definitely, I’m thinking about where to start, there’s so much to say about India. I’ll probably start from Echo devices, because it’s a bit more similar to what we have learned everywhere else. We started in a similar way where we built a natural authentic experience that we built in English. That was a bit limited, and then we added Hindi, and then we made the multilingual mode available. But even that doesn’t fully reflect the complexity of how customers interact with Alexa in general.

In India, the interaction is in multiple languages, so much deeper and complex that the switch (language) happens in the same phrase, or sentence. We need to go beyond the idea of speaking once in Hindi and once in English, we also need to recognise that many customers are also interacting in a lot more languages. Today, Alexa recognises basic commands in multiple languages, in part because it learned from customers.

Cricket and Bollywood content continue to drive engagement for Alexa in India
Photo Credit: Amazon

 

Since the beginning, we tried to create content that was engaging, such as deep knowledge of cricket and Bollywood. There are slightly unique expectations in behaviour in countries like India. We made services like music available in India right out of the box, unlike other countries where you have to have a different service provider. A big amount of music consumption is the newly released titles, driven primarily by Bollywood music. This involves updating the catalogue more frequently to be ready to play something that has just been released.

We also experimented with adding Alexa to the Amazon app, adding the convenience of voice interaction and Alexa intelligence to the Amazon app. Customers were able to ask for recommendations, add products to the cart, and other features. We are encouraged by the fact that some of the fastest-growing interactions involve shopping, even though people continue to use other use cases such as creating lists, setting alarms, etc.

One of the unique features we developed for India was Pay with Alexa. The payment industry in India is way more dynamic and exciting in India than anywhere else. Amazon Pay is a very interesting initiative, and adding the convenience of voice makes it easier than the touch experience.

Alexa drives all the voice-operated features on Amazon devices including Fire TV Cube
Photo Credit: Amazon

 

Gadgets 360: What are some of the biggest challenges you face when developing for markets like India?

Michele Butti: India is a bit unique; people speak a number of different languages. The first barrier was indeed understanding and that’s why we have made changes to our model to understand some of the commands in multiple languages. You can ask to play music in around 10-12 languages in India. The other challenge is the availability of broadband Internet, this is still not unique to India though. It will require broader industry development.

Gadgets 360: Do you plan to expand support for more regional languages in India?

Michele Butti: Probably the first way to answer your question is that we always listen to customers, you know when they’re passionate about something. We have clearly learned that even customers who are perfectly fluent in English, might have more engagement when English isn’t their preferred language. To add support to a language, it doesn’t just involve plain translation. It is the culture, the appreciation of preferences, history, and what is appropriate and what is not, that’s something that can change with region and language. This is where LLMs (large language models) can be an accelerator to help reduce the effort.

Gadgets 360: That means Alexa is getting generative AI support?

Michele Butti: Yes, we are already working on it. It is already training more languages than what we have made available so far.

Gadgets 360: At what point can Alexa become so smart that it can literally train itself?

Michele Butti: I think there are probably multiple components to this question. It starts from data labeling and transcription; we have already been able to dramatically reduce the amount of transcription needed to train the models. There is already supervised learning applied. LLM applications will likely help to further speed things up. AI will allow Alexa to react to more complex questions that were previously trained. There are use cases where LLMs (large language models) can make things easier, such as asking Alexa to record a basketball game (it will involve multiple steps and decision-making).

Gadgets 360: What’s the next big thing for Alexa?

Michele Butti: I don’t think there’s a single silver bullet. I’m personally very excited about making Alexa available in multiple smart properties. We mentioned hospitality, but it could be healthcare facilities or senior citizen facilities. We have launched some of those, that’s one other area where we’re making Alexa available. We have also launched Alexa in real estate, and smart properties. That way real estate developers can integrate Alexa in smart homes right from the beginning. Alexa is a big simplifier for companies that manage such large properties.

 

Disclosure: Amazon sponsored the correspondent’s flights and accommodation for the event in Seattle, US.