Akshat Kasera’s Post

🚨 New Tool Alert 🚨 Are you using ChatGPT to review your sales contracts? It works but it's risky. ChatGPT hallucinates a LOT. My friends at Outlit are building AI agents to review, revise, and approve your sales agreements in minutes. Key Features: ⚡ Fast AI Reviews: Get contracts reviewed in minutes with our AI agents. 🔍 Smart Search: Find any term, price, or clause quickly across all your contracts. 📘 Playbooks: Create guidelines to ensure contracts are reviewed the way you want. Check it out and show your support (upvotes, comments and feedback is welcome :) Link in the comments below: 👇

Friendly reminder: ChatGPT has limitations - they are: ☑ Biases - the model presents stereotypes or misinformation. ☑ Hallucinations - the model confidently states incorrect information. ☑ Overfitting - the model is only as good as the data it is trained on. If you rely on ChatGPT, be aware of these limitations. While it can be a powerful tool, It's crucial to verify the information it provides. Always cross-check with reliable sources to ensure accuracy. Remember: Use AI to augment creativity, not to replace it!

ChatGPT came first. But you cannot miss what’s next. AI agents. Explained best by Andrew Ng (see video). The rundown? ChatGPT = single prompt, single output LLMs today are primarily used in a non-agentic workflow. Prompt provided. Answer generated. Then your job is to follow up. What’s next?  Agentic Workflows (AI Agents). Agentic AI is iterative and can think and revise its output. More effective. Superior performance. Way more powerful. ………………………………………………. AI Agents at play: 1. MultiOn: Single AI agent tool. Learn a task and complete it autonomously. Example: Schedule a meeting. 2. HiveGPT: Many AI agents Agents multi task and collaborate to achieve goal. Example: Full mkt campaign in minutes not months.  The flow: - Upload creative brief. - Many AI agents get to work. - Build entire mkt campaign- content, landing pages, events. AI Agents= individuals, teams, humanity, empowered!  P.S. Are you looking forward to future of AI?

AI won't help your business stand out Here is the thing: ChatGPT is not smart. It knows what to say based on what other people have said before. It chooses the next words and phrases based on their probability of appearing after the last words. It is not in its DNA to be creative (yet). It is not even coded to be correct. It is coded to make smart guesses based on past information. The most correct information tends to be the average - so ChatGPT will always give you the average. The algorithm can be fine-tuned to be more "creative." This means it can be told to stray a bit from the middle line, but it must stay between two limits for the sentences to make sense. For this reason, it is hard for ChatGPT (or any other AI) to write a copy that will make your business stand out. It is not in its DNA to stand out but to blend in. AI, what is it good for? AI cannot develop original ideas but can help you better explain them. Feed it what you want to say, to whom, when, and all that jazz, and it will help you come up with the words your target audience uses. It can check your texts to make sure they make sense and are well-written. It can help you bounce ideas and come up with better examples—examples the average member of your audience will understand. AI won't make your business stand out, but your own ideas will.

A standard ChatGPT prompt represents less than 1% of AI capability. The true capability lies within the use of AI agent(s) - but what are they? Put simply, imagine turning the output of your ChatGPT prompt into a series of tasks which are performed either autonomously or with supervision. Now go one step further. Imagine if the input doesn't always follow the same structure. Or the output is dependent on a decision that needs to be made based on the input. This is the world of AI agents. Where an agent can ingest information and make a decision on the best course of action. And those courses of action could be a multitude of different processes and tasks. Simple example - if you work in sales and receive 100 emails a day, the chances are you are repeating 4-5 of the same processes depending on the contents of the email. Book a call, generate a proposal, hand over to the operations team, hand over to finance for invoicing etc. AI agents do this for you. They can interpret an email and run processes and tasks based on the semantic meaning of the email. And they do it automatically - 24/7 if needed. This framing is important to understand how AI & automation can work within your business. Processes in the past required specific inputs for specific outputs, but controlling specific inputs is tough in a world which continues to change. With AI agents, we no longer need to worry about this. And if used correctly the efficiency gains can be game-changing.

ChatGPT came first. But you cannot miss what’s next. AI agents. Explained best by Andrew Ng (see video). The rundown? ChatGPT = single prompt, single output LLMs today are primarily used in a non-agentic workflow. Prompt provided. Answer generated. Then your job is to follow up. What’s next? Agentic Workflows (AI Agents). Agentic AI is iterative and can think and revise its output. More effective. Superior performance. Way more powerful. ………………………………………………. AI Agents at play: 1. 𝗠𝘂𝗹𝘁𝗶𝗢𝗻: 𝗦𝗶𝗻𝗴𝗹𝗲 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁 𝘁𝗼𝗼𝗹. Learn a task and complete it autonomously. Example: Schedule a meeting. 2. 𝗛𝗶𝘃𝗲𝗚𝗣𝗧: 𝗠𝘂𝗹𝘁𝗶 𝗮𝗴𝗲𝗻𝘁 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 Agents multi-task and collaborate to achieve goal. Example: Full mkt campaign in minutes not months. The flow: - Upload creative brief. - Many AI agents get to work. - Build entire mkt campaign- content, landing pages, events. AI Agents= individuals, teams, humanity, empowered! P.S. Are you looking forward to future of AI?

The connection between AI and humans?🧠🤖 What concerns and considerations do technologies like ChatGPT raise for both employees and customers? 🤔🤔 When you query ChatGPT, a popular artificial intelligence (AI) tool, the responses are almost immediate, exude confidence, but are frequently incorrect.🤯🤯 🫣 It feels similar to conversing with an economist. The questions brought up by technologies like ChatGPT yield far more cautious answers. However, these are questions that managers should begin considering.😅 One concern is addressing employees' fears regarding job security. Such apprehensions are understandable.🤦🤦♀️ 😎An AI that simplifies expense processing is one thing; an AI that people would rather have as a dinner party companion is quite another.😎 Clearly outlining how workers can reallocate the time and effort saved by AI can help in gaining acceptance.🙌🙌

ChatGPT, Claude, and Perplexity Experience Downtime 🤖 Do you know why? In an unprecedented event, major AI chatbots including ChatGPT, Claude, and Perplexity all experienced outages simultaneously, leaving users scrambling for alternatives This incident has spotlighted the growing competition and reliability issues within the AI chatbot ecosystem During the downtime, many users turned to Claude, developed by Anthropic, and Perplexity AI. Claude has been noted for its sophisticated problem-solving capabilities and nuanced understanding, often outperforming ChatGPT in specific tasks like text analysis and detailed explanations (Tech.co) (Zeo Agency) Perplexity AI, meanwhile, leverages its strong search engine roots to deliver factual, concise responses with linked search results, offering a seamless blend of conversational AI and information retrieval (Zeo Agency). This makes it particularly appealing to users looking for quick, reliable facts The simultaneous outages of these leading AI chatbots have sparked discussions about the resilience and reliability of AI technologies, that in order to be successfull all the time need the support of #humans This is why KELONY has created a unique human-based #AI Tool, ARD Want to know more? DM us!

WHY YOU SHOULD USE CHATGPT AS POSSIBLE AS YOU CAN Innovation is dynamics process, holding past data does not make you improve rather than use the old data with new technology to create invisible skilled knowledge for the future. We will get new and improved knowledge and skills to create more value. And further more, I found that AI LLM is speeding up the process of turning explicit knowledge into tacit knowledge and this shows that people who can use AI tools is getting more valuble than who don't. Do you think data is important? Nope, I don't think so for now. Old data is past, even it holds value, but the new data will improve and make you fit the further and generate more value for you.

4 hacks that helped me master ChatGPT AI has the potential to level the playing field—𝗯𝘂𝘁 𝗺𝗼𝘀𝘁 𝗽𝗲𝗼𝗽𝗹𝗲 𝗮𝗿𝗲𝗻’𝘁 𝘂𝘀𝗶𝗻𝗴 𝗶𝘁 𝗲𝗳𝗳𝗲𝗰𝘁𝗶𝘃𝗲𝗹𝘆. Remember the movie 𝘓𝘪𝘮𝘪𝘵𝘭𝘦𝘴𝘴 with Bradley Cooper? We're there. Start simple and focus on the following: 𝟭. 𝗣𝗿𝗼𝗺𝗽𝘁 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 Craft clear, specific prompts to get high-quality responses. 𝟮. 𝗣𝗿𝗼𝘃𝗶𝗱𝗲 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 The more relevant background you give, the better the accuracy. 𝟯. 𝗜𝘁𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗥𝗲𝗳𝗶𝗻𝗲𝗺𝗲𝗻𝘁 Tweak your prompts, rephrase, and build on past responses for improvement. 𝟰. 𝗖𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 Always fact-check, challenge biases, and assess relevance. Rewatching Limitless last night, I remembered how incredible it seemed to achieve such cognitive ability. Fast-forward a decade, and here we are. Mastering AI is crucial for every aspiring entrepreneur—ignore it, and you’ll fall behind. I'll drop a great prompt in the comments to help you test founder/market fit today. DM me for more prompts!