Judhajit Sanyal’s Post

OpenAI promised a reasoning AI, & #chatGPT-4o1 just came. Next in their roadmap is #agency. What are the implications? Here’s a model that makes reasonable predictions relevant to 𝐨𝐮𝐫 𝐥𝐢𝐯𝐞𝐬 𝐚𝐧𝐝 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬. I developed a model for AI predictions that worked well so far. Here's a little piece of it. Natural Intelligence (𝐍𝐈) & especially Human Intelligence (𝐇𝐈) have: 𝐀𝐠𝐞𝐧𝐜𝐲: Capacity to make autonomous decisions and goals and act on them 𝐂𝐮𝐫𝐢𝐨𝐬𝐢𝐭𝐲: Drive to explore and seek new information and experiences for no reason 𝐀𝐦𝐛𝐢𝐭𝐢𝐨𝐧: Desire to achieve specific and challenging goals 𝐅𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐚𝐥 𝐂𝐮𝐫𝐢𝐨𝐬𝐢𝐭𝐲: goal-directed exploration 𝐄𝐱𝐞𝐜𝐮𝐭𝐢𝐨𝐧: Process and capacity to carry out decisions and achieve results 𝐁𝐨𝐬𝐬: Most NI and practically all HI have external entities that direct or limit their actions. The equivalent for AI is the 𝐀𝐈 𝐨𝐩𝐞𝐫𝐚𝐭𝐨𝐫. It’s a model, which means it is a simplified version of reality. Keep that in mind. My dog, cat, and parakeets all have those in various degrees. But let’s use the human standard as the reference point. Humans are 100% in all these areas. #AI4Leaders #AI #AiEvolution 𝟔 𝐩𝐚𝐠𝐞𝐬 / 𝟐 𝐦𝐢𝐧𝐮𝐭𝐞 𝐫𝐞𝐚𝐝

💡 #AI Quirks: A Fun Experiment with GPTs🚀 Recently, I was testing one known issue with different AI models, and something unexpected happened. #ChatGPT and #Claude both gave incorrect answers to a simple question—and the explanation that followed wasn’t helpful either (Claude's response was especially odd!). However, #Titan and #Gemini managed to get it right, but Titan couldn't explain why. It turns out the issue stems from how these models interpret words. For example, GPT sees the word ‘strawberry’ as three separate #tokens instead of individual letters, so it doesn't actually count the letters but instead predicts the answer based on patterns in its training data. This small experiment shows how each AI has strengths and weaknesses, and why it's important to use different models depending on the task. It's fascinating to see how subtle differences in processing can lead to vastly different outcomes!🔍🤖 Here comes #PromptEngineering into the picture, if we ask GPT's same question in different way(Chain of Thought) all gave correct answer: "Count how many R's are in the word strawberry. First, list each letter and indicate whether it's an r and tally as you go, and then give a count at the end." #AI #MachineLearning #GPT #COT #Claude #Titan #AIExperiments #Innovation

What is AI, how does it work and what can it be used for? 'Artificial intelligence (AI) technology is developing at high speed, transforming many aspects of modern life. There seem to be new announcements almost every day, with big players such as Meta, Google and ChatGPT-maker OpenAI competing to get an edge with customers.' #AI #ChatGPT #ArtificialIntelligence

Ever catch yourself relying on #AI for stuff you used to handle with no sweat? For example, I often find myself giving #ChatGPT sentences to correct for grammar that I could have easily done myself. Not even paragraphs, mere sentences. That's the Hidden Threat of AI: How It's Rewiring Our Decision-Making DNA 🔍 While we often worry about AI taking over, the real challenge is how it's sneaking into our decision-making process. These decisions, rooted in understanding, consideration of multiple options, and delayed closure, are essential for sound judgment. AI gives quick fixes, but it can skip the deep thinking we need for smart choices. Relying too much on AI means missing chances to sharpen our critical thinking skills. Let's find the right balance with AI, so we can still make decisions that match who we are and what we believe in. 💡 #ThoughtfulDecisions #AI #CriticialThinking

Artificial intelligence is now available to everyone. It isn’t just some far-flung sci-fi film idea any more. Should we be worried? Yes, I think so. Chat GPT is going to change the way we work forever, much like Google and the internet did. A.I. has opened up a new window of opportunity and those that use it to its advantage rather than hide their head in the sand will see great will have the upper hand. Are you worried about AI or are you already seeing its advantages? Let us know what you think in the comments. #chatgpt #artificialintelligence #businesstechnology #businessnewsdaily #ai #businesstools #businesstoolsandtips

I think it is inaccurate to say that Responsible AI is all about censorship and control. It is rather a standard setting exercise to ensure that systems provide the greatest good to people and society, ensuring that they feel included, safe, and their rights understood, accounted for and protected. It aims to create legitimate, reliable and accountable solutions to issues, rather than exacerbating existing problems or supporting the rapid creation of new ones. At its core is regulation of a technology that is developing at breakneck speed. So yes, maybe a part of it includes censorship to protect the rights of a larger user base, especially vulnerable groups like children, differently-able, and minorities, while exerting some control over tech firms that develop and promote these solutions. Other parts of this exercise would include adopting ethically focused approaches when designing, developing and executing AI systems, fostering collaborations and partnerships aimed at knowledge and experience sharing pertaining to new solutions, and investing time and money in education and dialogue on these areas to create a cultural, institutional, and systemic change to better inform the development of good technologies. This is not the first time technology has been regulated, and the need for responsibility emphasised.

Data has played an important role to help making informed decisions and reduce risks. I think the next big thing is going to be a promt library or templates which can help to generate accurate and close to expectations results to reduce additional work required on generated ai results. #AI #PromptEngineering #ChatGpt

While enjoying my afternoon snack—a very nice marine salad—let’s talk about AI or artificial intelligence. Just like salt in food, AI has become the must-have keyword to insert into any paragraph to gain more attention and possibly more value. This isn’t bad at all, as long as the keyword is used properly and in the correct context. However, the majority often confuse artificial intelligence with anything new or anything good. We see excessive use of the term when talking about LLMs (the GPT family of tools), which is only a tiny part of what artificial intelligence encompasses. Behind the scenes, these tools are simply algorithms processing matrices of possibilities to form good combinations based on a machine learning model. True intelligence has a more challenging meaning and deeper capabilities: learning from experience. ChatGPT will be intelligent when it can update its own LLM with the experience and interactions with users and the outside environment, such as interpreting world news, learning about events, and collecting global data. Without that, it remains a static LLM with a huge number of parameters and possibilities within its predefined matrix. #ai #gpt #artificial #intelligence #artificialintelligence #news #meaning

OpenAI on brink of breakthrough with "reasoning" AI, progress framework revealed. #AIprogress 🤝 Download 1 Million Logo Prompt Generator 🔜 https://wapia.in/1mlogo 🤝 Follow us on Whatsapp 🔜 https://wapia.in/wabeta _ ❇️ Summary: OpenAI has introduced a five-tier system to measure progress towards developing artificial general intelligence (AGI). The system ranges from basic chatbots to AI capable of managing entire organizations. The company believes its technology is currently at Level 1, with Level 2, "Reasoners," on the horizon. This classification system is still a work in progress and aims to provide a framework for understanding AI advancement. However, there is skepticism within the AI research community about the feasibility of achieving AGI and the accuracy of measuring progress towards it. Hashtags: #chatGPT AIbreakthrough ReasoningAI

AI News of the Day Stay updated with the latest in AI and tech developments. Here’s a quick glimpse: + OpenAI Dev Day! + AI Innovations! + Big Investments! P.S. Enjoyed this update? Feel free to share it with your audience ♻️ Have an amazing week! #linkedin #news #ai #artificialintelligence #chatgpt #technology