Ross McLean’s Post

Looking for AI use-cases — Benedict Evans https://buff.ly/3UiZbTY We’ve had ChatGPT for 18 months, but what’s it for? What are the use-cases? Why isn’t it useful for everyone, right now? Do Large Language Models become universal tools that can do ‘any’ task, or do we wrap them in single-purpose apps, and build thousands of new companies around that?Source: Looking for AI use-cases — Benedict Evans ()

Looking for AI use-cases — Benedict Evans https://buff.ly/3UiZbTY We’ve had ChatGPT for 18 months, but what’s it for? What are the use-cases? Why isn’t it useful for everyone, right now? Do Large Language Models become universal tools that can do ‘any’ task, or do we wrap them in single-purpose apps, and build thousands of new companies around that?Source: Looking for AI use-cases — Benedict Evans ()

Benedict Evans explores the practical applications and potential of Large Language Models (LLMs) like ChatGPT, emphasizing the challenges and opportunities they present. Despite the excitement around these models, Evans points out that many people, including himself, struggle to find everyday use-cases for them. He draws parallels to the early days of personal computing, where the utility of innovations like spreadsheets took time to become apparent and were initially relevant only to specific professions. Evans highlights that while AI tools have demonstrated significant potential in areas such as coding and generating first drafts, they often lack broad applicability and reliability in more complex, multi-step tasks. This is partly due to current limitations in AI accuracy and the probabilistic nature of these models, which can result in errors and "hallucinations." He suggests that for LLMs to reach their full potential, they must transition from being general-purpose tools to more specialized applications that address specific problems. This shift mirrors how previous technological innovations evolved, requiring new ideas and workflows to make the technology accessible and useful for different users. Evans concludes that the real challenge lies in identifying and developing these specific use-cases, which will involve significant creativity and effort from developers and entrepreneurs. This process will likely result in a proliferation of specialized applications, each leveraging the power of LLMs to automate tasks that were previously impossible or impractical to address with traditional software. https://lnkd.in/dFrX5TYQ

Here is how to justify spending $20 a month for the pro version of ChatGPT. Calculate what you’re worth an hour, and know that GPT-4 will save you at least 5 hours in your first month, especially in analyzing your data. It’s a no-brainer. If you’ve only used the free version of ChatGPT or are proudly telling people you’ve never paid for any AI, and you’ve been “fine,” you really don’t know what you’re missing. Trust me on this one. Yes, the free version can write you an ‘OK’ email, poem, or social media post. But GPT-4 can do those things ten times better PLUS browse the internet, generate images of anything you want, analyze and describe photos, have voice conversations with you, analyze all of your data and spreadsheets, make graphs from your data, and tons of other things. Try it for one month, and you’ll never be able to go back to 3.5 outputs. There's no comparison.

Yesterday was a great day for AI Operations Here's why: The next generation of businesses will operate differently. As we approach ~1 year of ChatGPT, it's only gotten clearer and clearer. Old teams were - Buried in busywork - Constantly chasing for full context - Limited by what they could execute, not what they thought should be done New teams will - Plan with full context - Delegate execution - Spend more time deciding what to do, not doing it If you haven't gotten serious about learning to delegate your work to AI. To build bespoke AI "systems" that carry out your work, the way you want it done, yesterday was a BIG sign you should be. OpenAI's custom GPT builder will make it so that ANY person, ANY company, can have their own bespoke AI assistants. You don't need to code. You just need to learn to do it (or work with people who can). The new goal of work isn't doing it; it's delegating it. -- BTW - If you're stuck anywhere in your internal AI adoption, me and my team would love to help. Shoot me a dm :)

A few cool things that have helped me integrate AI into my workflow this month: 📚 What is ChatGPT Doing and Why does it work by Stephen Wolfram - I keep going back to this to try and understand how this is all possible. Long, and there's some paragraphs that you have to just sit with for a while to really appreciate. 🤗 Huggingface.co - Easiest way to build and tweak AI models and see what's happening in the community. I think the platform itself will be really cool to watch since you have the data, models and hosting all wrapped up in one place. Ollama.ai is neat too. Obsidian Copilot - Run AI commands in note taking. Setting up custom prompts to format or rewrite notes. Pure AI generated text is still pretty trash and annoying, but using it to enhance or transform is really cool. GPT's - Instead of using chat, try providing instructions and building GPT's around specific problems. I'm working on one that's trained on my Kanban board for ShopFields to see how well it can build out a product board for a new POS app. I've got another one that's just trained on my car engine manual so ic an ask it questions on how that all works (mixed results)   I know I'm like a year late on most of these things - any other deep dive resources people recommend?

I just found out about this! 🧠 🎁 🤖 Save time and streamline your conversations with GPT-4 by using its “memory” feature . . . 1. In GPT-4, click on “Personal account” > “Settings” > “Personalization” 2. Toggle the “Memory” feature “on” > close the pop-up window 3. Within your prompt, tell GPT what to remember: “Always remember that when I ask for examples, I would like you to share 5 with me. Thanks!” 4. ChatGPT will indicate “Memory updated” = the request is stored in a memory library To access the memory library: 1. “Personal account” > “Settings” > “Personalization” > “Manage” 2. The memory library houses the user’s memorization requests 3. To remove an individual memory, click the trashcan icon associated with it 4. To delete all the Library’s contents, use the “Clear ChatGPT’s memory” button As The Neuron Daily newsletter (04/3024) explains: “Why it matters: Memory slightly outperforms Custom Instructions, but we view it as a mostly stopgap measure. It gets us one step closer to personalized AI assistants that truly remember everything, and I mean everything, about us: what projects we’re working on, preferred software, our brand voice, and more.” SOURCE: https://lnkd.in/gmWPcjWq

Did you think about: 1) If AI tech (like ChatGPT) will impact your work or business? 2) How will the impact look like? 3) How about indirect cross region impact? (ex. outsourcing/services demand changes) Feel free to drop a comment... #ai #business #economy #competition #artificialintelligence #chatgpt #machinelearning #work #business

𝗜𝘀 𝗚𝗣𝗧 𝗣𝗹𝘂𝘀 𝗦𝘁𝗶𝗹𝗹 𝗪𝗼𝗿𝘁𝗵 𝗜𝘁? OpenAI just released GPT-4o, and guess what? It's 𝗙𝗥𝗘𝗘 for everyone! The “o” is GPT-4o stands for “𝗢𝗺𝗻𝗶𝗰𝗵𝗮𝗻𝗻𝗲𝗹” because this latest AI model can understand text, audio, and images ALL AT ONCE!!! It's super fast and will be available for more people to use. But if you have ChatGPT Plus, you still get some exclusive perks. 𝗛𝗲𝗿𝗲’𝘀 𝘄𝗵𝗮𝘁 𝘆𝗼𝘂 𝗻𝗲𝗲𝗱 𝘁𝗼 𝗸𝗻𝗼𝘄: ⤷ With Plus, you can use GPT-4o way more before hitting limits or downgrading. ⤷ Free users get GPT-4 level access but Plus users get 5x more messages for GPT-4o. ⤷ Plus subscribers also get first dibs on cool new features like voice mode. ⤷ And if you're a Mac user, Plus gets you the slick desktop app too. ⤷ So, if you're a power user ChatGPT Plus is still worth the $20/month for all those extras. ˋˏ✄┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈ If you are interested in other cool stories like this, 𝗦𝗨𝗕𝗦𝗖𝗥𝗜𝗕𝗘 to our newsletter ai + u, it’s 𝗙𝗥𝗘𝗘! ⤷ https://aiplusyou.ai/

It isn’t a matter of how an “AI Model” will replace your professional services firm. It’s more of a matter when this will happen and those who use it in thier business will replace those who don’t. Starting from high in the sky from sales calls (Air), to proposals (Ignition) to responding to difficult emails (Compose AI) helping review data (Keeper.app), project management (ClickUp) presenting to clients (Canva), and most importantly use a note take app during a virtual meeting (Otter.ai.) which can help the client, staff or you recall important information. Yes, even with signing documents with Docusign & Adobe assisting in what parts to sign. If an AI model can learn how to create a story, it will learn how to do general accounting principles. For firm owners and those who are running professional service firms. Get the help you need by joining a mastermind, follow Jason Staats, CPA, join a community of entrepreneurs wanting to grow their business Sia Kal and most importantly do not do it alone. I can help and so many others can too. Be a student of your craft. #Learn #growing #business