Latest news with #AIcommunity
Forbes
a day ago
- Forbes
Here Are GPT-5 Prompt Engineering Insights Including Crucial AI Prompting Tips And Techniques
In today's column, I provide GPT-5 prompt engineering tips and techniques that will aid in getting the best outcomes when using this newly released generative AI. I'm sure that just about everyone by now knows that OpenAI finally released GPT-5, doing so after a prolonged period of immense and wildly fantastical speculation about what it would be like. Well, now we know what it is (see my in-depth review of GPT-5 at the link here). Bottom line is that GPT-5 is pretty much akin to all the other generative AI and large language models (LLMs) when it comes to doing prompting. The key is that if you want to ensure that GPT-5 works suitably for your needs, you must closely understand how GPT-5 differs from prior OpenAI AI products. GPT-5 has distinctive features and functionality that bring forth new considerations about composing your prompts. Let's talk about it. This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here). Readers might recall that I previously posted an in-depth depiction of over eighty prompt engineering techniques and methods (see the link here). Top-notch prompt engineers realize that learning a wide array of researched and proven prompting techniques is the best way to get the most out of generative AI. Prompting Is Still Tried And True The first place to begin when assessing GPT-5 from a prompt engineering perspective is that prompts are still prompts. Boom, drop the mic. I say that somewhat irreverently. Here's the deal. There was prior conjecture that perhaps GPT-5 would turn the world upside down when it came to using prompts. The floated ideas of how GPT-5 might conceivably function were astounding and nearly out of this world ('it will read your mind', 'it will know what you want before you even know', etc.). The truth is now known. GPT-5 is essentially a step-up from ChatGPT and GPT-4, but otherwise you do prompting just like you've done all along. There isn't a new kind of magical way to write prompts. You are still wise to compose prompts as you've been doing since the early days of contemporary generative AI. To clarify, I am emphasizing that you should astutely continue to write clearly worded prompts. Be direct. Don't be tricky. Write prompts that are long enough to articulate your question or task at hand. Be succinct if possible. Definitely don't be overly profuse or attempt to be complicated in whatever your request is. And so on. Those are all golden rules and remain perfectly intact when using GPT-5. I am confident that all the prompt engineering specialized techniques that I've previously covered will generally work appropriately with GPT-5. Some might require a tweak or minor refinement, but otherwise, they are prudent and ready to go (see my list at the link here). Auto-Switching Can Be A Headache We can next consider how to artfully try and accommodate GPT-5 via composing prompts that GPT-5 will efficiently and effectively act on. The biggest aspect that entails both good news and bad news about GPT-5 is that OpenAI decided to include an auto-switcher. This is a doozy. It will require you to potentially rethink some of your prompting since it is quite possible that GPT-5 isn't going to make the right routing decisions on your behalf. Allow me a moment to explain the quandary. It used to be that you would have to choose which of the various OpenAI AI products you wanted to use for a particular situation at hand. There had been an organic expansion of OpenAI's prior models in the sense that there have been GPT-4o, GPT-4o-mini, OpenAI o3, OpenAI o4-mini, GPT-4.1.-nano, and so on. When you wanted to use OpenAI's AI capabilities, you had to select which of those available models you wanted to utilize. It all depended on what you were looking to do. Some were faster, some were slower. Some were deeper at certain classes of problems, others were shallower. It was a smorgasbord that required you to pick the right one as suitable for your task at hand. The onus was on you to know which of the models were particularly applicable to whatever you were trying to do. It could be a veritable hit-and-miss process of selection and tryouts. GPT-5 now has uplifted those prior versions into new GPT-5 submodels, and the overarching GPT-5 model makes the choice of which GPT-5 submodel might be best for whatever problem or question you happen to ask. The good news is that depending on how your prompts are worded, there is a solid chance that GPT-5 will select one of the GPT-5 submodels that will do a bang-up job of answering your prompt. The bad news is that the GPT-5 auto-switcher might choose a less appropriate GPT-5 submodel. Oops, your answer will not be as sound as if the more appropriate submodel had been chosen. Worse still, each time that you enter a prompt or start a new conversation, the GPT-5 auto-switcher might switch you to some other GPT-5 submodel, back and forth, doing so in a wanton fashion. It can make your head spin since the answers potentially will vary dramatically. Craziness In Design The average user probably won't realize that all these switcheroo mechanics are happening behind the scenes. I say that because GPT-5 doesn't overtly tell you that it is taking these actions. It just silently does so. I appreciate that the designers apparently assumed that no one would care or want to know what is going on under the hood. The problem is that those who are versed in using AI and are up-to-speed on prompting are being bamboozled by this hidden and secreted behavior. A savvy user can almost immediately sense that something is amiss. Frustratingly, GPT-5 won't let you directly control the auto-switching. You cannot tell the AI to use a particular submodel. You cannot get a straight answer if you ask GPT-5 which submodel it intends to use on your prompt. It is perhaps like trying to get the key to Fort Knox. GPT-5 refuses to play ball. The marketplace has tweeted vociferously that something needs to be done about this lack of candor by GPT-5 regarding the model routing that is occurring. Sam Altman sent out a tweet on X that suggested they are going to be making some changes on this aspect (see his X posting of August 8, 2025). The thing is, we can applaud the desire to have a seamless, unified experience, but it is similar to having an automatic transmission on a car. Some users are fine with an automatic transmission, but other, more seasoned drivers want to know what gear the car is in and be able to select a gear that they think is most suitable for their needs. Prompting GPT-5 For Routing As the bearer of bad news, I should also add that the auto-switching comes with another said-to-be handy internal mechanism that decides how much processing time will be undertaken for your entered prompt. Again, you have no particular say in this. It could be that the prompt gets tons of useful processing time, or maybe the time is shortchanged. You can't especially control this, and the settings are not within your grasp (as an aside, to some degree, if you are a developer and are using the API, you have more leeway in dealing with this; see the OpenAI GPT-5 System Card for the technical details). Let me show you what I've been doing about this exasperating situation. First, here is a mapping of the prior models to the GPT-5 submodels: The GPT-5 submodels are considered successors and depart from the earlier models in various ways. That being said, they still are roughly on par as to the relative strengths and weaknesses that previously prevailed. I will show you what I've come up with to try and sway the GPT-5 auto-switcher. Prompting With Aplomb Suppose I have a prompt that I believe would have worked best on GPT-4o. But I am using GPT-5, thus I am not using GPT-4o, plus OpenAI has indicated that it will sunset the prior models, so you might as well get used to using GPT-5. Darned if you cannot simply tell GPT-5 to use gpt-5-main (i.e., realizing that gpt-5-main is now somewhat comparable to GPT-4o, per my chart above). The AI will either tell you it doesn't function that way or might even imply that it will do as you ask, yet it might do something else. Bow to the power of the grand auto-switcher. This eerily reminds me of The Matrix. Anyway, we need to somehow convince GPT-5 to do what we want, but we must do so with aplomb. Asking straightaway isn't a viable option. The need to sway the AI is our best option at this ugly juncture. In the specific case of my wanting to use gpt-5-main, here is a prompt that I use and seems to do the trick (much of the time): It appears that by emphasizing the nature of what I want GPT-5 to do, it seems possible to sway the direction that the auto-switcher will route my next prompt. Not only will I possibly get the submodel that I think is the best choice for the prompt, observe that I also made a big deal about the depth of reasoning that ought to take place. This potentially helps to kick the AI into giving an allotment of processing time that it, by enigmatic means, would have perhaps inadvertently shortcut (OpenAI refers to processing time as so-called 'thinking time' – an anthropomorphizing of AI that I find to be desperate and despairing). I am not saying this sway-related prompting is a guaranteed result. After trying a bunch of times, it seemed to be working as hoped for. I came up with similar prompts for each of the other GPT-5 submodels. If there is enough interest expressed by readers, I will do a follow-up with those details. Be on the watch for that upcoming coverage. On a related note, I will also soon be covering the official GPT-5 Prompting Guide that OpenAI has posted, along with their Prompt Optimizer Tool. Those are aimed primarily at AI developers and not especially about day-to-day, ordinary prompting in GPT-5. Watch Out That Writing Is Enhanced On the writing side of things, GPT-5 has improvements in a myriad of writing aspects. The ability to generate poems is enhanced. Depth of writing and the AI being able to make more compelling stories and narratives seems to be an added plus. My guess is that the everyday user won't discern much of a difference. For a more seasoned user, you are bound to notice that the writing has gotten an upgrade. I suppose it is something like getting used to a third grader and now being conversational with a sixth grader. Or something like that. I use this prompt to get GPT-5 to be closer to the way it was in the GPT-4 series: That seems to get me the kind of results that I used to see. It is not an ironclad method, but it generally works well. I realize that some people are going to scream loudly that I ought not to suggest that users revert to the GPT-4 writing style. We all should accept and relish the GPT-5 writing style. Are we going backwards by asking for GPT-5 to speak like GPT-4? Maybe. I grasp the angst. It's up to you, and I'm not at all saying that everyone should use this prompting tip. Please use it at your personal discretion. Lies And AI Hallucinations OpenAI claims that GPT-5 is more honest than prior OpenAI models, plus it is less likely to hallucinate (hallucination is yet another misappropriated word used in the AI field to describe when the AI produces fictionalized responses that have no bearing in fact or truth). I suppose it might come as a shock to some people that AI has been and continues to lie to us, see my discussion at the link here. I would assume that many people have heard or even witnessed that AI can make things up, i.e., produce an AI hallucination. Worries are that AI hallucinations are so convincing in their appearance of realism, and the AI has an aura of confidence and rightness, that people are misled into believing false statements and, at times, embrace its crazy assertions. See more at the link here. A presumed upbeat consideration is that apparently GPT-5 reduces the lying and reduces the AI hallucinations. The downbeat news is that it isn't zero. In other words, it is still going to lie and still going to hallucinate. This might happen on a less frequent basis, but nonetheless remains a chancy concern. Here is my prompt to help try and further reduce the odds of GPT-5 lying to you: Here is my prompt to help further reduce the odds of GPT-5 incurring a so-called hallucination: My usual caveats apply, namely, these aren't surefire, but they seem to be useful. The crucial motto, as always, still is that if you use generative AI, make sure to remain wary and alert. One other aspect is that you would be shrewd to use both of those prompts so that you can simultaneously try to strike down the lying and the hallucinations. If you only use one of those prompts, the other unresolved side will potentially arise. Try to squelch both. It's your way of steering out of the range of double trouble. Personas Are Coming To The Fore I've repeatedly emphasized in my writing and talks about generative AI that one of the most underutilized and least known pieces of quite useful functionality is the capability of forming personas in the AI (see the link here). You can tell the AI to pretend to be a known person, such as a celebrity or historical figure, and the AI will attempt to do so. For example, you might tell AI to pretend to be Abraham Lincoln. The AI will respond based on having pattern-matched on the writings of Lincoln and the writings about Lincoln. It is instructive and useful for students and learners. I even showcased how telling AI to simulate Sigmund Freud can be a useful learning tool for mental health professionals, see the link here. OpenAI has indicated they are selectively making available a set of four new preset personas, consisting of Cynic, Robot, Listener, and Nerd. Each of those personas represents those names. The AI shifts into a mode reflecting those types of personalities. The good news is that I hope this spurs people to realize that personas are a built-in functionality and easily activated via a simple prompt. It doesn't take much work to invoke a persona. Here is my overall prompt to get a persona going in GPT-5: Use personas with due caution. I mention this because some people kind of get lost in a conversation where the AI is pretending to be someone. It isn't real. You aren't somehow tapping into the soul of that actual person, dead or alive. Personas are pretenses, so keep a clear head accordingly. Prompt Engineering Still Lives I hope that these important prompting tips and insights will boost your results when using GPT-5. One last comment for now. You might know that some have fervently claimed that prompt engineering is a dying art. No one will need to write prompts anymore. I've discussed in great depth the automated prompting tools that try to do the prompting for you (see my aforementioned list of prompt engineering strategies and tactics). They are good and getting better, but we are still immersed in the handwriting of prompts and will continue down this path for quite a while to come. GPT-5 abundantly reveals that to be the case. A final remark for now. It has been said that Mark Twain made a wry comment that when a newspaper reported him as deceased, he said that the audacious claim was a bit exaggerated. That was smarmily tongue-in-cheek. I would absolutely say the same about prompt engineering. It's here. It isn't disappearing. Keep learning about prompting. You'll be glad that you spent the prized time doing so.
Geeky Gadgets
30-07-2025
- Geeky Gadgets
New Claude AI Usage Rate Limits : How Will You Adapt to the New Restrictions?
What if you woke up tomorrow to find your most trusted AI assistant suddenly limited in how much it could help you? For a small but significant group of users—just 5%—this scenario is now a reality. Anthropic's recent introduction of new weekly usage limits for its Claude AI plans has sparked a wave of questions, concerns, and adaptations. While these changes aim to promote fairness and sustainability, they also raise critical challenges for heavy users who depend on Claude for complex workflows. If you're someone who's built your productivity or creativity around this tool, these updates could feel like a seismic shift, forcing you to rethink how you work with AI. The question is: are you in the 5%, and if so, how will you adapt? In this overview of the new Claude usage limits, Ray Fernando breaks down what these new limits mean for users and explore the ripple effects they're creating across the AI community. You'll gain a clearer understanding of the updated caps, why they've been implemented, and how they might affect your workflows. Whether you're a developer managing intricate codebases or a creator pushing the boundaries of AI-driven projects, this shift could redefine how you approach your work. We'll also uncover strategies to optimize your usage, weigh the pros and cons of alternative solutions, and examine the broader implications for ethical AI practices. As you read on, consider how these changes might not just challenge your current processes but also inspire new ways to innovate within the constraints. New Claude Usage Limits What Are the New Weekly Usage Limits? The updated plans establish specific usage caps based on the service tier: Pro Users 40-80 hours of Sonnet 4 through Claude Code per week Max Plan ($100/month) 140-280 hours of Sonnet 4 per week 15-35 hours of Opus 4 per week Max Plan ($200/month) 240-480 hours of Sonnet 4 per week 24-40 hours of Opus 4 per week These limits primarily affect heavy users, such as developers managing extensive codebases or running multiple parallel sessions. If your usage patterns align with these scenarios, you may encounter these thresholds sooner than anticipated. For users exceeding the limits, Anthropic provides the option to purchase additional usage at standard API rates. While this offers flexibility, it also raises questions about cost-effectiveness for those with high computational demands. Evaluating your workflows and resource allocation will be critical to navigating these changes efficiently. Why Are These Changes Being Made? The implementation of these limits reflects Anthropic's commitment to addressing misuse and promoting fair usage across its platform. Common policy violations that prompted these changes include: Sharing accounts among multiple users Reselling AI services through unauthorized proxies Engaging in continuous 24/7 usage patterns that strain system resources By enforcing these restrictions, Anthropic aims to ensure equitable resource distribution and prevent exploitation of the service. If you've relied on practices such as shared accounts or other unauthorized methods to maximize access, adapting to these rules will be necessary to remain compliant. These measures also reflect a broader effort to maintain the platform's integrity and sustainability as its user base continues to grow. New Claude Rate Limits : Are You in the 5%? Watch this video on YouTube. Here are more detailed guides and articles that you may find helpful on Claude AI. How to Monitor and Optimize Your Usage To help users stay within the new limits, Anthropic recommends using token consumption monitoring tools such as `bunx CC usage` or `npx CC usage`. These tools provide real-time insights into your usage patterns, allowing you to: Track token consumption: Gain a clear understanding of how resources are being used. Gain a clear understanding of how resources are being used. Identify inefficiencies: Pinpoint areas in your workflows where resource usage can be optimized. Pinpoint areas in your workflows where resource usage can be optimized. Allocate resources effectively: Avoid exceeding limits by adjusting your processes proactively. This is particularly valuable for developers managing complex projects or running resource-intensive tasks. Proactive monitoring not only helps you stay compliant but also ensures that your workflows remain uninterrupted under the new constraints. Claude 7 Day Rate Limits Explained Watch this video on YouTube. Here are some tips to maximize your usage: Start new conversations for each topic. Batch related questions together. Be mindful of attachment sizes as they impact usage limits. Use projects to store documents in project knowledge instead of repeatedly uploading them. Break large documents into smaller sections before uploading. Ask Claude to summarize key points that you can paste into a new chat when needed. Regularly review and remove unnecessary text and images from your project knowledge to keep your context focused and relevant. Options for Heavy Users For users who find the new limits restrictive, exploring alternative solutions may be necessary. Some viable options include: Open source AI models: These models offer greater flexibility and control over resources. However, they may require additional setup, maintenance, and technical expertise to integrate into your workflows effectively. These models offer greater flexibility and control over resources. However, they may require additional setup, maintenance, and technical expertise to integrate into your workflows effectively. Sub-agent delegation: Distributing workloads across multiple accounts or agents can help reduce the strain on individual accounts. This approach requires careful planning to ensure seamless integration with existing processes. While these strategies can provide additional freedom, they also come with their own challenges. Evaluating the trade-offs between flexibility, cost, and complexity will be essential to determine the best path forward for your specific needs. Community Feedback and Broader Implications The announcement of these changes has sparked significant discussion within the developer community. Key concerns raised include: Transparency: Users are calling for clearer guidelines on how usage is tracked and monitored. Users are calling for clearer guidelines on how usage is tracked and monitored. Workflow impact: Many developers are concerned about how the new limits will affect their productivity and project timelines. Many developers are concerned about how the new limits will affect their productivity and project timelines. Cost implications: Heavy users worry about the financial burden of purchasing additional usage or transitioning to alternative solutions. These changes have also prompted broader conversations about ethical AI usage and resource sustainability. Balancing productivity, cost, and fairness is a challenge that extends beyond individual users, highlighting the need for industry-wide solutions. As AI technologies continue to evolve, fostering open dialogue and collaboration within the community will be critical to addressing these challenges effectively. Anthropic's Perspective From Anthropic's viewpoint, these updates are a necessary step to ensure the long-term sustainability of their services. By addressing policy violations and promoting fair usage, the company aims to maintain high-quality service for the majority of users while managing the demands of a rapidly growing user base. While the changes may pose challenges for heavy users, they reflect a broader commitment to ethical and responsible AI usage. Anthropic's approach underscores the importance of balancing resource availability with equitable access, making sure that the platform remains viable for all users in the long term. Adapting to these updates will be essential for those seeking to maximize the benefits of Claude AI while aligning with the platform's evolving policies. Media Credit: Ray Fernando Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
Geeky Gadgets
17-07-2025
- Science
- Geeky Gadgets
Context Rot and Overload : How Feeding AI More Data Can Decrease Accuracy
What happens when the very thing designed to make AI smarter—more context—starts to work against it? Large Language Models (LLMs), celebrated for their ability to process vast amounts of text, face a surprising Achilles' heel: as input lengths grow, their performance often falters. This phenomenon, sometimes referred to as 'context rot,' reveals a paradox at the heart of AI: the more we feed these models, the harder it becomes for them to deliver accurate, coherent results. Imagine asking an AI to summarize a 50-page report, only to receive a response riddled with errors or omissions. The culprit isn't just the complexity of the task—it's the overwhelming flood of tokens that dilutes the model's focus and reliability. In this feature, Chroma explore the intricate relationship between input length and LLM performance, uncovering why even the most advanced models struggle with long and complex inputs. From the role of ambiguity and distractors to the uneven prioritization of context, you'll gain a deeper understanding of the hidden challenges that affect AI's ability to reason and respond. But it's not all bad news—by adopting strategies like summarization and retrieval, users can mitigate these issues and unlock more consistent, reliable outputs. How can we strike the right balance between context and clarity? Let's delve into the mechanics of context rot and the solutions that can help us harness the true potential of LLMs. How Long Inputs Impact LLM Performance When tasked with processing long inputs, LLMs often experience a noticeable decline in both accuracy and efficiency. While they excel at handling short and straightforward tasks, their performance diminishes when faced with extended or intricate inputs. This issue becomes particularly evident in tasks requiring reasoning or memory, where the model must integrate information from multiple sections of the input. For instance, even seemingly simple tasks like replicating structured data or answering direct questions can become unreliable when the input includes excessive or irrelevant context. The sheer volume of tokens can overwhelm the model, leading to errors, inconsistencies, and a reduced ability to focus on the primary task. These limitations highlight the importance of tailoring inputs to the model's capabilities to maintain performance and reliability. The Role of Ambiguity and Distractors Ambiguity and distractors are two critical factors that exacerbate the challenges LLMs face when processing long inputs. These elements can significantly impair the model's ability to generate accurate and relevant outputs. Ambiguity: When input questions or content lack clarity, the model may interpret multiple possible meanings. This often results in outputs that are either incorrect, overly generalized, or fail to address the intended query. When input questions or content lack clarity, the model may interpret multiple possible meanings. This often results in outputs that are either incorrect, overly generalized, or fail to address the intended query. Distractors: Irrelevant or misleading information embedded within the input can divert the model's attention from the core task. This reduces the accuracy and reliability of the output, as the model struggles to differentiate between critical and non-essential details. To mitigate these issues, it is essential to structure inputs carefully and eliminate unnecessary or confusing elements. By doing so, you can help the model maintain focus and precision, making sure more consistent and reliable results. AI Context Rot : How Increasing Input Tokens Impacts LLM Performance Watch this video on YouTube. Advance your skills in AI performance by reading more of our detailed content. Inconsistent Context Processing Another significant limitation of LLMs is their inconsistent handling of long inputs. As the context window fills, the model may prioritize certain sections of the input while neglecting others. This uneven processing can lead to incomplete or irrelevant outputs, particularly for tasks that require a comprehensive understanding of the entire input. For example, summarizing a lengthy document or extracting key details from a complex dataset becomes increasingly error-prone as the input grows. The model's inability to uniformly process all parts of the context undermines its reliability in such scenarios. This inconsistency highlights the need for strategies that ensure the model can focus on the most relevant portions of the input without losing sight of the broader context. Strategies for Effective Context Management To address these challenges, adopting robust context management strategies is essential. Two primary approaches—summarization and retrieval—have proven effective in optimizing LLM performance when dealing with long inputs: Summarization: Condensing lengthy inputs into shorter, more relevant summaries reduces the cognitive load on the model. By retaining only the most critical information, summarization improves the accuracy and relevance of the outputs while minimizing the risk of errors caused by excessive or irrelevant context. Condensing lengthy inputs into shorter, more relevant summaries reduces the cognitive load on the model. By retaining only the most critical information, summarization improves the accuracy and relevance of the outputs while minimizing the risk of errors caused by excessive or irrelevant context. Retrieval: Using vector databases to fetch only the most pertinent information for a given task can significantly enhance performance. This involves indexing input data into a searchable format, allowing the model to access specific pieces of information as needed, rather than processing the entire input at once. In many cases, combining these strategies can yield even better results. For example, summarization can be used to refine the input before retrieval, making sure that the model focuses solely on the most relevant details. This layered approach allows for more efficient and accurate processing of complex or lengthy inputs. Key Takeaways Despite advancements in token capacities and context window sizes, LLMs continue to face challenges when processing long inputs. Issues such as ambiguity, distractors, and inconsistent context handling can hinder their performance and reliability. However, by employing effective context management strategies like summarization and retrieval, you can optimize the model's ability to generate accurate and relevant outputs. Understanding these limitations and tailoring your approach to specific use cases is essential for using the full potential of LLMs. By structuring inputs carefully and adopting strategic context management techniques, you can overcome the challenges of context rot and ensure more consistent and reliable results from these powerful tools. Media Credit: Chroma Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
Geeky Gadgets
14-07-2025
- Business
- Geeky Gadgets
Grok 4 vs ChatGPT : Exploring the Future of Advanced AI Reasoning
What if the tools we rely on to solve our most complex problems could think more deeply, reason more effectively, and collaborate more intelligently than ever before? Enter Grok 4, a innovative AI model that promises to redefine advanced reasoning. Positioned as a bold leap forward in artificial intelligence, Grok 4 is not just another incremental update—it's a statement about the future of problem-solving. Yet, as with any ambitious innovation, it comes with its own set of challenges and trade-offs. Can this new model truly deliver on its promise, or does it reveal the limits of AI's current capabilities? In this exploration, Matt Wolfe unpacks the fantastic potential of Grok 4, examining how it tackles complex analytical tasks and where it falls short. Along the way, we'll also delve into the broader AI landscape, from AI-powered browsers to new advancements in video technology and healthcare. These innovations are not just reshaping industries; they're raising critical questions about ethics, accessibility, and the future of collaboration between humans and machines. As we navigate this rapidly evolving terrain, one thing becomes clear: the race to harness AI's full potential is as thrilling as it is uncertain. AI Advancements and Trends Grok 4: Advanced Reasoning at the Forefront Grok 4 represents a leap forward in AI reasoning, designed to tackle complex analytical challenges with precision. It offers tiered pricing plans tailored to different user needs, starting at $30/month for the 'Super Grok' plan and scaling up to $300/month for the 'Grok 4 Heavy' plan. The 'Heavy' plan is particularly notable for its use of multiple agents, allowing deeper insights and more sophisticated analysis. This makes it a powerful tool for professionals and organizations requiring advanced problem-solving capabilities. Despite its strengths, Grok 4 has certain limitations. It struggles with real-time information retrieval, which can hinder its effectiveness in fast-paced environments. Additionally, it lacks the extensive integrations offered by competitors like ChatGPT, which provide broader functionality. However, for users prioritizing high-level reasoning over general versatility, Grok 4 remains a compelling option. AI-Powered Browsing: Perplexity's Comet Browser The Comet browser by Perplexity introduces a new paradigm in web navigation, blending AI capabilities with a Chrome-based platform. This browser integrates an AI assistant to streamline everyday tasks, such as comparing prices, managing emails, and summarizing webpages. By automating these processes, Comet enhances productivity and reduces the manual effort required for routine activities. Currently, the browser is available to Perplexity Max subscribers at $200/month and is being rolled out on an invite-only basis. This exclusivity reflects the growing competition in the AI browser market, with rumors suggesting that OpenAI is developing its own browser to challenge Comet and other platforms. As AI-driven browsing tools continue to evolve, they promise to redefine how users interact with the web. Grok 4 vs ChatGPT Watch this video on YouTube. Explore further guides and articles from our vast library that you may find relevant to your interests in Advanced reasoning models. Innovations in AI Video Technology AI is transforming video creation, offering tools that expand creative possibilities and streamline production workflows. V3's image-to-video feature is a standout innovation, allowing consistent character animations with synchronized audio. This tool is particularly valuable for content creators aiming to produce high-quality visual narratives efficiently. Meanwhile, Moon Valley's ethical AI video model emphasizes responsible innovation. Trained exclusively on licensed data, it offers advanced customization options, such as motion and pose transfer, while adhering to strict ethical standards. These developments highlight the importance of balancing creativity with accountability, making sure that AI-driven video technology is both powerful and responsible. Collaborative AI Systems and Industry Dynamics Collaboration between AI systems is becoming increasingly sophisticated, allowing new levels of efficiency and problem-solving. Sakana AI's TreeQuest model exemplifies this trend by allowing multiple AI agents to work together on complex tasks. This collaborative approach not only enhances productivity but also broadens the scope of AI applications across industries. At the same time, the AI industry is witnessing intense competition for top talent. Companies like Meta and OpenAI are aggressively recruiting professionals, with Meta attracting talent from Apple and OpenAI targeting experts from Tesla and XAI. These recruitment efforts underscore the high stakes involved in shaping the future of AI and the race to achieve new advancements. AI in Healthcare and Beyond The influence of AI extends far beyond traditional tech applications, driving significant breakthroughs in fields like healthcare. For example, Google's Isomorphic Labs is advancing AI-designed drugs to human trials, offering new hope for patients and researchers. These developments have the potential to transform medicine by accelerating drug discovery and improving treatment outcomes. In infertility treatments, AI is being used to identify more effective solutions, showcasing its ability to address some of the most pressing challenges in science and medicine. These applications demonstrate how AI is not only transforming industries but also improving lives on a global scale. AI-Driven Monetization and Crypto Integration The integration of AI into monetization strategies and financial platforms is reshaping how value is created and exchanged. YouTube, for instance, has updated its monetization policies to address concerns about repetitive or mass-produced AI-generated content. This move reflects a broader effort to ensure quality and originality in AI-driven media. Simultaneously, Perplexity's partnership with Coinbase introduces crypto data into its platform, highlighting the growing intersection of AI and blockchain technologies. These initiatives illustrate how AI is being used to drive innovation and create new opportunities across diverse sectors, from content creation to financial services. Speculation on AGI and the Future of AI As AI continues to advance, speculation about the timeline for achieving Artificial General Intelligence (AGI) is intensifying. The aggressive recruitment strategies of leading companies suggest a race to unlock AGI's fantastic potential. While the exact timeline remains uncertain, the pursuit of AGI underscores the profound possibilities that lie ahead for AI and its role in shaping the future. The rapid pace of innovation in AI is reshaping industries and redefining possibilities. From Grok 4's advanced reasoning capabilities to Perplexity's AI-powered browser and breakthroughs in video technology, these developments illustrate the fantastic potential of AI. As competition intensifies and new applications emerge, staying informed about these advancements is essential for navigating the evolving landscape and harnessing the opportunities it presents. Media Credit: Matt Wolfe Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
Forbes
12-07-2025
- Science
- Forbes
United Nations Considering These Four Crucial Actions To Save The World From Dire AGI And Killer AI Superintelligence
The United Nations releases an important report on AGI and emphasizes four key recommendations to ... More help save the world from dire outcomes. In today's column, I examine a recently released high-priority report by the United Nations that emphasizes what must be done to prepare for the advent of artificial general intelligence (AGI). Be aware that the United Nations has had an ongoing interest in how AI is advancing and what kinds of international multilateral arrangements and collaborations ought to be taking place (see my coverage at the link here). The distinctive element of this latest report is that the focus right now needs to be on our reaching AGI, a pinnacle type of AI. Many in the AI community assert that we are already nearing the cusp of AGI and, in turn, we will soon thereafter arrive at artificial superintelligence (ASI). For the sake of humanity and global survival, the U.N. seeks to have a say in the governance and control of AGI and ultimately ASI. Let's talk about it. This analysis of an innovative AI breakthrough is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here). Heading Toward AGI And ASI First, some fundamentals are required to set the stage for this weighty discussion. There is a great deal of research going on to further advance AI. The general goal is to either reach artificial general intelligence (AGI) or maybe even the outstretched possibility of achieving artificial superintelligence (ASI). AGI is AI that is considered on par with human intellect and can seemingly match our intelligence. ASI is AI that has gone beyond human intellect and would be superior in many if not all feasible ways. The idea is that ASI would be able to run circles around humans by outthinking us at every turn. For more details on the nature of conventional AI versus AGI and ASI, see my analysis at the link here. We have not yet attained AGI. In fact, it is unknown whether we will reach AGI, or whether AGI may be achievable in decades or perhaps centuries from now. The AGI attainment dates that are floating around are wildly varying and wildly unsubstantiated by any credible evidence or ironclad logic. ASI is even more beyond the pale when it comes to where we are currently with conventional AI. United Nations Is Into AI And AGI I've previously explored numerous U.N. efforts regarding where AI is heading and how society should best utilize advanced AI. For example, I extensively laid out the ways that the U.N. recommends that AI be leveraged to attain the vaunted Sustainability Development Goals (SDGs), see the link here. Another important document by the U.N. is the UNESCO-led agreement on the ethics of AI, which was the first-ever global consensus involving 193 countries on the suitable use of advanced AI (see my analysis at the link here) The latest notable report is entitled 'Governance of the Transition to Artificial General Intelligence (AGI): Urgent Considerations for the UN General Assembly' and was prepared and submitted to the Council of Presidents of the United Nations General Assembly (UNCPGA). Here are some key points in that report (excerpts): The bottom line is that a strong case can be made that if AGI is allowed to be let loose and insufficiently overseen, society is going to be at grave risk. A question arises as to how the nations of the world can unite to try and mitigate that risk. Aptly, the United Nations believes they are the appropriate body to take on that challenge. UN Given Four Big Asks What does the U.N. report say about urgently needed steps regarding coping with the advent of AGI? These four crucial recommendations are stridently called for: Those recommendations will be considered by the Council of Presidents of the United Nations General Assembly. By and large, enacting one or more of those recommendations would indubitably involve some form of U.N. General Assembly resolutions and would undoubtedly need to be integrated into other AI initiatives of the United Nations. It is possible that none of the recommendations will proceed. Likewise, the recommendations might be revised or reconstructed and employed in other ways. I'll keep you posted as the valued matter progresses. Meanwhile, let's do a bit of unpacking on those four recommendations. I will do so, one by one, and then provide a provocative or perhaps engaging conclusion. Global AI Observatory The first of the four recommendations entails establishing a global AGI Observatory that would keep track of what's happening with AGI. Think of this as a specialized online repository that would serve as a curated source of information about AGI. I agree that this would potentially be immensely helpful to the U.N. Member States, along with being useful for the public at large. You see, the problem right now is that there is a tremendous amount of misinformation and disinformation concerning AGI that is being spread around, often wildly hyping or at times undervaluing the advent of AGI and ASI. Assuming that the AGI Observatory was properly devised and suitably careful in what is collected and shared, having a source about AGI that is reliable and balanced would be quite useful. One potential criticism of such an AGI Observatory would be that it is perhaps duplicative of other similar commercial or national collections about AGI. Another qualm would be if the AGI Observatory were allowed to be biased, it would misleadingly carry the aura of something balanced, yet would actually be tilted in a directed way. Best Practices And Certification For AGI The second recommendation requests that a set of AGI best practices be crafted. This would aid nations in understanding what kind of governance structures ought to be considered for sensibly overseeing AGI in their respective country. It could spur nations to proceed on a level playing field basis. Furthermore, it reduces the proverbial reinventing of the wheel, namely that the nations could simply adopt or adapt an already presented set of AGI best practices. No need to write such stipulations from scratch. On a similar vein, the setting up of certifications for AGI would be well-aligned with the AGI best practices. AI makers and countries as a whole would hopefully prize being certified as to their AGI and its conformance to vital standards. A criticism on this front is that if the U.N. does not make the use of best practices a compulsory aspect, and likewise if the AGI certification is merely optional, few if any countries would go to the trouble of adopting them. In that sense, the whole contrivance is mainly window dressing and not a feet-to-the-fire consideration. U.N. Framework Convention In the parlance of the United Nations, it is somewhat expected to call for a Framework Convention on significant topics. Since AGI is abundantly a significant topic, here's a snapshot excerpt of what is proposed in the report: 'A Framework Convention on AGI is needed to establish shared objectives and flexible protocols to manage AGI risks and ensure equitable global benefit distribution. It should define clear risk tiers requiring proportionate international action, from standard-setting and licensing regimes to joint research facilities for higher-risk AGI, and red lines or tripwires on AGI development.' The usual criticism of those kinds of activities is that they can become a bureaucratic nightmare that doesn't produce much of anything substantive. Also, they might stretch out and be a lengthy affair. This is especially disconcerting in this instance if you believe that AGI is on the near horizon. Formulate U.N. AGI Agency The fourth recommendation indicates that a feasibility study be undertaken to assess whether a new U.N. agency ought to be set up. This would be a specialized U.N. agency devoted to the topic of AGI. The report stresses that this would need to be quickly explored, approved, and set in motion on an expedited basis. An analogous type of agency or entity would be the International Atomic Energy Agency (IAEA). You probably know that the IAEA seeks to guide the world toward peaceful uses of nuclear energy. It has a founding treaty that provides self-guidance within the IAEA. Overall, the IAEA reports to the U.N. General Assembly and the U.N. Security Council. A criticism of putting forward an AGI Agency by the United Nations is that it might get bogged down in international squabbling. There is also a possibility that it would be an inhibitor to the creative use of AGI rather than merely serving as a risk-reducing guide. To clarify, there are some that argue against too many regulating and overseeing bodies since this might undercut innovative uses of AGI. We might inadvertently turn AGI into something a lot less impressive and valuable than we had earlier hoped for. Sad face. Taking Action Versus Sitting Around Do you think that we should be taking overt governance action about AGI, such as the recommendations articulated in the U.N. AGI report? Some would say that yes, we must act immediately. Others would suggest we take our sweet time. Better to get things right than rush them along. Still others might say there isn't any need to do anything at all. Just wait and see. As food for thought on that thorny conundrum, here's a memorable quote by Albert Einstein: 'The world will not be destroyed by those who do evil, but by those who watch them without doing anything.' Mull that over and then make your decision on what we should do next about AGI and global governance issues. The fate of humanity is likely on the line.
