Sam Altman Blames Compute Scaling for Lack of GPT-5

Instead, Orion will be available only to the companies OpenAI works closely with. OpenAI has dropped a couple of key ChatGPT upgrades so far this year, but neither one was the big GPT-5 upgrade we’re all waiting for. First, we got GPT-4o in May 2024 with advanced multimodal support, including Advanced Voice Mode. Then more recently, we got o1 (in preview) with more advanced reasoning capabilities. GPT-5 is also expected to show higher levels of fairness and inclusion in the content it generates due to additional efforts put in by OpenAI to reduce biases in the language model.

• The company plans to regularly update and improve these models, including adding features like browsing, file and image uploading, and function calling, which are currently not available in the API version.
• Screenshots provided to Ars Technica found that ChatGPT is potentially leaking unpublished research papers, login credentials and private information from its users.
• All of these models have gotten quite complex and we can’t ship as many things in parallel as we’d like to.
• The company’s goal is to combine its LLMs over time to create an even more capable model that could eventually be called artificial general intelligence, or AGI.
• The next few months will be critical in determining whether GPT-5 can deliver on its promise of a significant leap forward, addressing the limitations of its predecessors and paving the way for more advanced AI applications.
• Given the talk of OpenAI pitching partnerships with publishers, the AI biz may be looking to show off how it can summarize current news content in its chatbot replies, which would be search-adjacent.

With that denial, the exact details on the rumored AI model have been tricky to pin down. However, an OpenAI executive has claimed that “Orion” aims to have 100 times more computation power than GPT-4. While the number of parameters in GPT-4 has not officially been released, estimates have ranged from 1.5 to 1.8 trillion. That means lesser reasoning abilities, more difficulties with complex topics, and other similar disadvantages.

Apple announced at WWDC 2024 that it is bringing ChatGPT to Siri and other first-party apps and capabilities across its operating systems. The ChatGPT integrations, powered by GPT-4o, will arrive on iOS 18, iPadOS 18 and macOS Sequoia later this year, and will be free without the need to create a ChatGPT or OpenAI account. Features exclusive to paying ChatGPT users will also be available through Apple devices.

The Buzz Around ‘Project Strawberry’

During a demonstration of ChatGPT Voice at the VivaTech conference, OpenAI’s Head of Developer Experience Romain Huet showed a slide revealing the potential growth of AI models over the coming few years and GPT-5 was not on it. When GPT-3 came out, the entire AI space—and the tech industry in general—reacted with shock. Many said it was revolutionary, and some immediately declared that it meant AGI was imminent.

The models are also available via the OpenAI API for developers who qualify for API usage tier 5, though initial rate limits will apply. Additionally, the o1-preview model excels in coding, ranking in the 89th percentile in Codeforces competitions, showcasing its ability to handle multi-step workflows, debug complex code, and generate accurate solutions. OpenAI envisions the models being used for a wide range of applications, from helping physicists generate mathematical formulas for quantum optics to assisting healthcare researchers in annotating cell sequencing data. Heller said he did expect the new model to have a significantly larger context window, which would allow it to tackle larger blocks of text at one time and better compare contracts or legal documents that might be hundreds of pages long. In November 2023 OpenAI’s board of directors ousted Altman from his role as CEO stating that he hadn’t been forthcoming in his communications with the board and they didn’t “trust him to lead” the company any longer. But in a dramatic reversal of fortune, Microsoft hired Altman — and a few other ex-OpenAI execs — three days later to run an advanced AI research project.

Few AI features and applications are truly unique, and only a handful are compelling enough to justify the AI PC label. Sure, AI PCs may have Neural Processing Units with some impressive performance, but outside of getting you better battery life and better hardware acceleration, there hasn’t been a “Killer App” for the AI market. These updates “had a much stronger response than we expected,” Altman told Bill Gates in January. To address these issues, the Microsoft-backed company is collaborating with Broadcom and TSMC to design its own chips aimed at boosting computing capacity. Altman confirmed that OpenAI does not plan to release the next major AI model, GPT-5, this year. 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.

The Verge also notes that Orion is seen as the successor of GPT-4, but it’s unclear if it’ll keep the GPT-4 moniker or tick up to GPT-5. GPT-5 will be more compatible with what’s known as the Internet of Things, where devices in the home and elsewhere are connected and share information. It should also help support the concept known as industry 5.0, where humans and machines operate interactively within the same workplace. Arthur has been a tech journalist ever since 2013, having written for multiple sites. He really got into tech when he got his first tablet, the Archos 5, back in 2011.

The new model brings with it improvements in writing, math, logical reasoning and coding, OpenAI claims, as well as a more up-to-date knowledge base. OpenAI has partnered with another news publisher in Europe, London’s Financial Times, that the company will be paying for content access. “Through the partnership, ChatGPT users will be able to see select attributed summaries, quotes and rich links to FT journalism in response to relevant queries,” the FT wrote in a press release. OpenAI planned to start rolling out its advanced Voice Mode feature to a small group of ChatGPT Plus users in late June, but it says lingering issues forced it to postpone the launch to July. OpenAI says Advanced Voice Mode might not launch for all ChatGPT Plus customers until the fall, depending on whether it meets certain internal safety and reliability checks. But the feature falls short as an effective replacement for virtual assistants.

NYT tech workers are making their own games while on strike

With the app, users can quickly call up ChatGPT by using the keyboard combination of Option + Space. The app allows users to upload files and other photos, as well as speak to ChatGPT from their desktop and search through their past conversations. After a big jump following the release of OpenAI’s new GPT-4o “omni” model, the mobile version of ChatGPT has now seen its biggest month of revenue yet. The app pulled in $28 million in net revenue from the App Store and Google Play in July, according to data provided by app intelligence firm Appfigures. OpenAI has found that GPT-4o, which powers the recently launched alpha of Advanced Voice Mode in ChatGPT, can behave in strange ways.

Equally, it can automatically create a new image that matches the user’s prompt, or text description. For instance, the system’s improved analytical capabilities will allow it to suggest possible medical conditions from symptoms described by the user. GPT-5 can process up to 50,000 words at a time, which is twice as many as GPT-4 can do, making it even better equipped to handle large documents. However, GPT-5 will have superior capabilities with different languages, making it possible for non-English speakers to communicate and interact with the system. The upgrade will also have an improved ability to interpret the context of dialogue and interpret the nuances of language. So, as time goes on, we can expect OpenAI to release fewer and fewer updates.

As it turns out, the GPT series is being leapfrogged for now by a whole new family of models. Heller’s biggest hope for GPT-5 is that it’ll be able to “take more agentic actions”; in other words, complete tasks that involve multiple complex steps without losing its way. This could include reading a legal fling, consulting the relevant statute, cross-referencing the case law, comparing ChatGPT it with the evidence, and then formulating a question for a deposition. Our community is about connecting people through open and thoughtful conversations. We want our readers to share their views and exchange ideas and facts in a safe space. Whether or not the Orion news is 100% accurate, it’s fair to note that it’s likely overshadowed by OpenAI’s seemingly unending news cycle.

More and more tech companies and search engines are utilizing the chatbot to automate text or quickly answer user questions/concerns. Aptly called ChatGPT Team, the new plan provides a dedicated workspace for teams of up to 149 people using ChatGPT as well as admin tools for team management. In addition to gaining access to GPT-4, GPT-4 with Vision and DALL-E3, ChatGPT Team lets teams build and share GPTs for their business needs. Premium ChatGPT users — customers paying for ChatGPT Plus, Team or Enterprise — can now use an updated and enhanced version of GPT-4 Turbo.

But still, Sam Altman’s vision of a super-competent AI colleague is both exciting and transformative. This AI would go beyond being a tool, becoming a true partner that enhances our abilities and enriches our lives. By providing deep knowledge, proactive assistance and creative collaboration, it could help us achieve more than we ever thought possible. As we move toward this future, addressing the challenges of privacy and bias will be essential to ensure that this advanced AI serves as a positive force in our lives.

“We have some very good releases coming later this year! Nothing that we are going to call gpt-5, though,” he said during a Reddit AMA this week. OpenAI CEO Sam Altman has poured cold water on hopes for the next major version of ChatGPT coming out this year. From Meta’s AI-empowered AR glasses to its new Natural Voice Interactions feature to Google’s AlphaChip breakthrough and ChromaLock’s chatbot-on-a-graphing calculator mod, this week has been packed with jaw-dropping developments in the AI space. He has not expounded on his position since publishing that tweet, leaving us confused both about his statement’s meaning and his company’s eventual plans for Orion.

In terms of its safety, Altman has posted on X (formerly Twitter) that OpenAI would be “working with the US AI Safety Institute,” and providing early access to the the next foundation model. According to the report from The Verge, Orion won’t actually release as a part of ChatGPT. Instead, it would reportedly be limited to partnerships with specific companies — at least at first. OpenAI and its peers can’t expect that everyone creating digital content will want to have their work included in an AI model that enriches model makers but not anyone else. And those whose work has already been incorporated into existing models may have something to say on the matter too, if the law allows it.

Mobile users are being pushed to upgrade to its $19.99 monthly subscription, ChatGPT Plus, if they want to experiment with OpenAI’s most recent launch. OpenAI is testing SearchGPT, a new AI search experience to compete with Google. SearchGPT aims to elevate search queries with “timely answers” from across the internet, as well as the ability to ask follow-up questions. The temporary prototype is currently only available to a small group of users and its publisher partners, like The Atlantic, for testing and feedback.

Whatever the timing, it’s clear that we’re fast approaching a release of something big from the market leader. When a new model comes out, it will get better at reasoning, it will perform better across all of the standard metrics and benchmarks, allowing for improved coding, better writing, and more nuanced conversations with AI. For example, we know for a fact that GPT-4.0 is capable of creating images, vector graphics, and the voice version is capable of singing, and all of these features have been disabled by OpenAI. I’ve written many stories speculating on new features coming to the next version of iOS, a button changing on the side of a pair of headphones, or a camera update in the latest smartphone.

The response, signed by CEO Sam Altman and Chairman of the Board Bret Taylor, said building a complete and diverse board was one of the company’s top priorities and that it was working with an executive search firm to assist it in finding talent. In an effort to win the trust of parents and policymakers, OpenAI announced it’s partnering with Common Sense Media to collaborate on AI guidelines and education materials for parents, educators and young adults. The organization works to identify and minimize tech harms to young people and previously flagged ChatGPT as lacking in transparency and privacy.

This includes the ability to make requests for deletion of AI-generated references about you. Although OpenAI notes it may not grant every request since it must balance privacy requests against freedom of expression “in accordance with applicable laws”. We will see how handling troubling statements produced by ChatGPT will play out over the next few months as tech and legal experts attempt to tackle the fastest moving target in the industry. ChatGPT is AI-powered and utilizes LLM technology to generate text after a prompt. Users will also be banned from creating chatbots that impersonate candidates or government institutions, and from using OpenAI tools to misrepresent the voting process or otherwise discourage voting. Beginning in February, Arizona State University will have full access to ChatGPT’s Enterprise tier, which the university plans to use to build a personalized AI tutor, develop AI avatars, bolster their prompt engineering course and more.

It enhanced the model’s ability to handle complex queries and maintain longer conversations, making interactions smoother and more natural. GPT-2 was like upgrading from a basic bicycle to a powerful sports car, showcasing AI’s potential to generate human-like text across various applications. Additionally, if OpenAI’s GPT models ever ChatGPT App achieve Artificial General Intelligence (AGI), the partnership between Microsoft and OpenAI will dissolve. This is clearly problematic for Microsoft, as OpenAI’s GPT technology is at the heart of Microsoft’s Copilot AI software platform. Other reports indicate that GPT-4o “Strawberry” and GPT-5 could cost $2,000 for users to run.

Here are a couple of features you might expect from this next-generation conversational AI. ChatGPT-5 is definitely coming with several groundbreaking features and enhancements that could level up how we interact with AI. Building on the success of GPT-3, ChatGPT-4 brought further refinements in understanding and generating text.

• “I think maybe AI is going to not super significantly but somewhat significantly change the way people use the internet,” Altman said.
• Instead, he now apparently thinks models will likely continue to grow, driven by significant investments in computing power and energy.
• For example, there are chatbots that are rules-based in the sense that they’ll give canned responses to questions.
• ChatGPT-5 will be better at learning from user interactions and fine-tuning its responses over time to become more accurate and relevant.

After being delayed in December, OpenAI plans to launch its GPT Store sometime in the coming week, according to an email viewed by TechCrunch. OpenAI says developers building GPTs will have to review the company’s updated usage policies and GPT brand guidelines to ensure their GPTs are compliant before they’re eligible for listing in the GPT Store. OpenAI’s update notably didn’t include any information on the expected monetization opportunities for developers listing their apps on the storefront. In a blog post, OpenAI announced price drops for GPT-3.5’s API, with input prices dropping to 50% and output by 25%, to $0.0005 per thousand tokens in, and $0.0015 per thousand tokens out. GPT-4 Turbo also got a new preview model for API use, which includes an interesting fix that aims to reduce “laziness” that users have experienced.

In theory, this additional training should grant GPT-5 better knowledge of complex or niche topics. It will hopefully also improve ChatGPT’s abilities in languages other than English. But a significant proportion of its training data is proprietary — that is, purchased or otherwise acquired from organizations. In practice, that could mean better contextual understanding, which in turn means responses that are more relevant to the question and the overall conversation. On the other hand, there’s really no limit to the number of issues that safety testing could expose. Delays necessitated by patching vulnerabilities and other security issues could push the release of GPT-5 well into 2025.

That’s a problem when you’re using it to do your homework, sure, but when it accuses you of a crime you didn’t commit, that may well at this point be libel. After some back and forth over the last few months, OpenAI’s GPT Store is finally here. The feature lives in a new tab in the ChatGPT web client, and includes a range of GPTs developed both by OpenAI’s partners and the wider dev community. As part of a test, OpenAI began rolling out new “memory” controls for a small portion of ChatGPT free and paid users, with a broader rollout to follow. You can foun additiona information about ai customer service and artificial intelligence and NLP. The controls let you tell ChatGPT explicitly to remember something, see what it remembers or turn off its memory altogether. Note that deleting a chat from chat history won’t erase ChatGPT’s or a custom GPT’s memories — you must delete the memory itself.

That’s because ChatGPT lacks context awareness — in other words, the generated code isn’t always appropriate for the specific context in which it’s being used. A chatbot can be any software/system that holds dialogue with you/a person but doesn’t necessarily have to be AI-powered. For example, there are chatbots that are rules-based in the sense that they’ll give canned responses to questions.

According to The Verge, engineers at Microsoft Azure, OpenAI’s cloud service provider, are getting ready to launch Orion on the Azure platform, potentially starting in November. To achieve this level of capability it needs to have all the abilities and skills of the previous stages plus broad intelligence. To run an organization it would need to be able to understand all the independent parts and how they work together. Level 4 is where the AI becomes more innovative and capable of “aiding in invention”. This could be where AI adds to the sum of human knowledge rather than simply draws from what has already been created or shared. Compare having a conversation with Siri or Alexa to that of ChatGPT or Gemini — it is night and day and this is because the latter is a conversational AI.

A Step Closer to AGIWhile the world eagerly awaits the launch of GPT-5, reports indicate that the AI model is likely to arrive no sooner than early 2025. There was speculation about a December 2024 release, but a company spokesperson denied those rumours, possibly due to recent leadership changes within OpenAI, including the departure of former CTO Mira Murati. Sam Altman further shared his intentions to open up the “Not Safe For Work” (NSFW) adult content sometime in the future and let users experience interactions with AI like adults without limitations. However, given the challenges such an implementation could pose, it is still a future consideration. He specifically said that he would not be releasing the GPT-5 this year and would instead focus on shipping GPT-o1.

However, based on the company’s past release schedule, we can make an educated guess. Efficiency improvements in ChatGPT-5 will likely result in faster response times and the ability to handle more simultaneous interactions. This will make the AI more scalable, allowing businesses and developers to deploy it in high-demand environments without compromising performance. This would open up a ton of new applications, such as assisting in video editing, creating detailed visual content, and providing more interactive and engaging user experiences. ChatGPT-5 is likely to integrate more advanced multimodal capabilities, enabling it to process and generate not just text but also images, audio, and possibly video. One of the most significant improvements expected with ChatGPT-5 is its enhanced ability to understand and maintain context over extended conversations.

Paid users of ChatGPT can now bring GPTs into a conversation by typing “@” and selecting a GPT from the list. The chosen GPT will have an understanding of the full conversation, and different GPTs can be “tagged in” for different use cases and needs. According to a report from The New Yorker, ChatGPT uses an estimated 17,000 times the amount of electricity than the average U.S. household to respond to roughly 200 million requests each day. According to Reuters, OpenAI’s Sam Altman hosted hundreds of executives from Fortune 500 companies across several cities in April, pitching versions of its AI services intended for corporate use. On the The TED AI Show podcast, former OpenAI board member Helen Toner revealed that the board did not know about ChatGPT until its launch in November 2022.

The less prevalent water is in a given region, and the less expensive electricity is, the more likely the data center is to rely on electrically powered air conditioning units instead. In Texas, for example, the chatbot only consumes an estimated 235 milliliters needed to generate one 100-word email. That same email drafted in Washington, on the other hand, would require 1,408 milliliters (nearly a liter and a half) per email. The nonprofit portion of the business will not be done away with entirely, but instead would continue to exist and own a minority stake in the overall company. So, while a launch later this December seems plausible, timed with the two-year anniversary of ChatGPT, it’s just as likely that it won’t come until 2025 based on how inaccurate all the predictions have been so far.

Orion is viewed internally as a successor to GPT-4, though it is unclear whether its official name will be GPT-5 when released. An OpenAI executive has reportedly hinted that Orion could be up to 100 times more powerful than GPT-4, Open AI’s flagship model. Regardless of what product names OpenAI chooses for future ChatGPT models, the next major update might be released by December. But this GPT-5 candidate, reportedly called Orion, might not be available to regular users like you and me, at least not initially. The technology behind these systems is known as a large language model (LLM).

With GPT-4 we saw a model with the first hints of multimodality and improved reasoning and everyone expected GPT-5 to follow the same path — but then a small team at OpenAI trained GPT-4o and everything changed. Nigel Powell is an author, columnist, and consultant with over 30 years of experience in the technology industry. He produced the weekly Don’t Panic technology column in the Sunday Times newspaper for 16 years and is the author of the Sunday Times book of Computer openai gpt-5 Answers, published by Harper Collins. He has been a technology pundit on Sky Television’s Global Village program and a regular contributor to BBC Radio Five’s Men’s Hour. The story, from The Verge, also suggests Microsoft is being given an inside track on this release, and could deliver a version for Azure sometime in November. If true, this would mark the first time that Microsoft has been given free rein to openly release such a major AI product before OpenAI itself.