AI shows no sign of slowing down

“It’s not perfect, but neither are you.” Those were some of the final words uttered by Greg Brockman, President of OpenAI, during yesterday’s live demo of GPT-4. The demo has now been viewed 1 million times. During the 25-minute video, Greg walked through several examples (mentioned below) of how GPT-4 can take your work to the next level. But the three things that stood out to me from the announcement are: an extra focus on ethics, it’s a work in progress, and it’s all about potential. Ethics, progress, potential. I view this overall release as less “game-changing” than the ChatGPT launch, I mean that’s a tough one to beat, but in the public evals and in my personal testing, there seems to be a large performance boost. Though I still sometimes prefer 3.5 for speed and length, I’m also cautiously optimistic about the much larger context window that the 32k model will provide (and can’t believe more people aren’t talking about it). The use cases laid out in the video are similar to ones I’ve shared previously, with one big exception: ✅ Summarization - submit a long paragraph or essay, get a recap ✅ Document comparison - submit two pieces of content, get the commonalities ✅ Coding - submit a coding task in nearly any language, get ready-made code ✅ Code debugging - submit new documentation and GPT-4 will fix outdated code ✅ Answers and calculations - submit complex documentation like taxes to get helpful explanations and walkthroughs The BIG exception, and where my eyes widened during the demo, was when Greg took a crappy sketch he drew of a fun website idea, and GPT-4 created code for the entire website…buttons and all…based on his notebook drawing. GPT-4 takes in both text AND image input, and that will crack open many more use cases. Caveats and limitations: 👎 They’re working on making it faster (in my testing, it’s definitely slower than previous models, but not enough to not use it) 👎 It’s not available to everyone just yet and the team asks for patience as they roll it out. Based on documentation I read, there is also limited availability of the 32k token model and a rate limit of 100 msgs per 4 hours to GPT-4 in ChatGPT Plus 👎 Data still cuts off September 2021 👎 Improved writing also means higher potential for disinformation and influenced extremism I’ll end with Greg’s quote toward the end of the demo, and why it’s important that everyone, not just AI experts, poke at these systems and understand their purposes and limitations, and use this technology (or not) to build the future they want. “This is all just potential. You can see lots of different applications. We, ourselves, are still figuring out new ways to use this.” What do you think? What questions do you still have? Follow Allie K. Miller for more on AI and tech. #gpt4 #openai #chatgpt #artificialintelligence #technology #future #work

Great article on the BBC about Chat GPT4 and AI in general and whether AI is and will be developed transparently revealing and policing where the large language models draw their information? Also, whether governments have the knowledge, experience, ability and desire to regulate AI and protect our data privacy or will they abrogate that responsibility? 2 great questions. The answers to both dictate whether AI makes life easier and more seamless, as AI features on consumer products can anticipate user needs and help accomplish virtually any task or involves less data privacy, more centralisation of power in a handful of companies and AI that anticipates user needs - and gets them wrong or, at least, stifles choices. #ai #dataprivacy #dataprotection https://lnkd.in/eGmkVeVr

Crazy about ChatGPT Plus with the new GPT-4? You should be. But craziness sometimes pushes for narrowed vision, so let us see what else happens in the AI Large Language Models ecosystem. Or in other words, let us take a look at #Claude: - Anthropic has launched a new chatbot, Claude, designed to rival OpenAI's ChatGPT. - The AI startup, founded by former OpenAI researchers, aims to make AI systems safer and more understandable. So briefly, comparing to OpenAI's products: - OpenAI's ChatGPT is based on gpt-3.5-turbo, while Anthropic has not disclosed the underlying model for Claude. - Both chatbots serve various applications such as drafting emails, writing code and answering questions. - Claude is reportedly worse as a programming and math assistant. Not to say about the higher rate of hallucinations (Source: https://lnkd.in/eXC8hstm ) Anticipated impact: - Claude powers DuckDuckGo’s recently launched DuckAssist tool. Quora offers access to Claude through its experimental AI chat app, Poe. And on Notion, Claude is a part of the technical backend for Notion AI. - Claude's entry into the AI market signifies growing competition for OpenAI. - Remember, Google invested $300 mln for a 10% stake at Anthropic, so yes, the heat is on them, but also this proves Google sees GPT tech as a massive threat. In short: - Anthropic's new chatbot, Claude, is a rival to OpenAI's ChatGPT, and its launch signifies growing competition in the AI chatbot market. The rivalry may drive further innovation and development of safer, more reliable AI systems. - Yet it lags behind core parts of the 'crazy offering' that GPT-4 and its interface: ChatGPT- Plus, brings to the table. Programming functionalities. I am very interested to see how the next months will work for these companies. #ai

Is the world ready for AI? 🌎💻🤖 I hear some concerns and apprehension about how it could impact our jobs, economy, and society as a whole. It's important that we continue to educate ourselves about AI and its capabilities, while also developing ethical guidelines and regulations to ensure its responsible use. 👩💻 Read more HERE: https://lnkd.in/eHDMGGgY #AI #tech #future #responsibility

As I thought would happen a month ago, after taking a tepid step with Bard, Google is throwing caution to the wind and opening its AI crown jewel PaLM to developers 24 hours after OpenAI rolled out GPT-4. Now we have a real race! I expect an open trial conversional AI from Google soon. This race will be over in a year. Every day counts. It’s a huge risk but GPT-4 is just that good. ________________________________________ Mobile/Retail/Home Platforms are transforming medical research and healthcare delivery! Exponential technologies like artificial intelligence, gene sequencing/multi-omics, and wearables are accelerating the transformation! I cover both in my LinkedIn posts and #NewHealthcare Insider Newsletter at https://lnkd.in/eMpq7X7m. [Opinions are personal and based on public information]

yesterday was the biggest AI launch day; GPT-4, Google, Anthropic, Adept, Intercom, Duolingo, Poe, Notion and so many more companies launched or announced something big our 3 picks of the day are: 1/ 🤯 GPT-4 is here Enter GPT-4, the long-awaited update to the model underlying ChatGPT! It has come with lots of ‘mind-blown’ emojis over its additional performance quality, disappointment about OpenAI’s lack of transparency, and the ability to involve both images and text - known as ‘multimodality’ According to OpenAI, GPT-4 has hit its best results yet on factuality, steerability, and keeping within guidelines It’s even taken ChatGPT from being in the 10th percentile to the 90th percentile on the bar, as well as drastic improvements on many other human exams Beyond the much-expected language improvements, it can generate responses to prompts and documents that include images - all whilst maintaining similar skills 2/ 🤖 Google drop their own generative AI tools, with inclusions in Google Docs As soon as OpenAI dropped the GPT-4 bomb, they dropped their own slick new AI developer tool! The PaLM API is their upcoming tool that provides a way to build apps on top of their best language models; and they’ve coupled it with MakerSuite, the sister tool to help with prototyping ideas and more They’ve also announced a bunch of new companies and tech partners’ training models with them, including Midjourney! But the coolest part of all of this is the inclusion of generative AI within their workspace tools a.k.a Google Docs and Gmail Type out a topic and it’ll generate a draft, allowing you to adjust for tone and style too Hold your horses though - it’s only available to ‘trusted’ developers, so you’ll have to wait! 3/ 💸 Adept get a $350m funding round for their laptop-assistant AI vision At Adept, their vision is centered around allowing AI to work with you on your computer, letting it develop with you and execute operations for you at your request This could allow your creativity and productivity to run wild The possibilities might be endless with a wad of cash that big #ai #ArtificialIntelligence #ChatGPT #Google #OpenAI

I wrote a new book with OpenAI's latest, most powerful large language model. It’s called Impromptu: Amplifying our Humanity through AI. This, as far as I know, is the first book written with GPT-4. Last summer, I got access to GPT-4. It felt like I had a new kind of passport. My pages were quickly filled with stamps: Over 1,000 prompts. 800+ pages of outputs. Just in the first few months. With GPT-4, I traveled through light bulb jokes, epic poems, original sci fi plots, arguments about human nature, musings on how AI might strengthen democracy, society and industries. The goal, like in any good trip, was to learn as much about my traveling partner as the place I was exploring. Earlier this year, I organized my explorations with GPT-4 into sections: creativity, education, criminal justice, public intellectualism, journalism and more. My travelog, Impromptu, captures my treks with GPT-4. These paths will be well-trodden soon (if not already) by others. I'm writing this travelog both to encourage people to not only get to know GPT-4, but to embrace our choice of how we’ll use it, and explore the different ways this choice might play out. Impromptu is now available: As a free PDF: https://lnkd.in/eZSsGYgB   On Kindle: https://lnkd.in/eUE2ukM3  Here’s a quick video on the origin and making of Impromptu:

The most interesting thing in tech: GPT-4 is here, and we’re about to enter a world where everything written is a blur—where we have great new powers but almost no idea who wrote what. #mostinterestingthingintech

OpenAI has just launched GPT-4, the company's most advanced system to date. With broader general knowledge and more accurate problem-solving abilities, GPT-4 can solve difficult problems more accurately. In addition, GPT-4 is more creative and collaborative than ever before, capable of generating, editing, and iterating with users on creative and technical writing tasks, such as composing music, writing scripts, or learning a user's writing style. Another new feature of GPT-4 is its ability to accept images as input and generate captions, classifications, and analyses. It can also handle more than 25,000 words of text, allowing for use cases such as long-form content creation, extended conversations, and document analysis and research. GPT-4 also surpasses ChatGPT in its advanced reasoning and performance capabilities. To ensure the safety and alignment of GPT-4, OpenAI spent 6 months working on improvements and testing the model internally. Although GPT-4 has known limitations such as social biases, hallucinations, and adversarial prompts, OpenAI is working to address them and encourage transparency and AI literacy. GPT-4 was trained on Microsoft Azure AI supercomputers and is available on the ChatGPT Plus platform, with APIs soon to be released for developers to build applications and services. Tell me your thoughts in the comments below Follow me for more interesting updates #openai #chatgpt #gpt #gpt4 #gpt3 #microsoft #inteligenciaartificial #testing #creative #learning #ai #datamusingwithdisha

🔥 The tech world is currently under the AI buzz as ChatGPT has successfully made everyone's life a bit easier. Now that #ChatGPT is supported by Microsoft and Bing, how are the other tech giants responding?🧐 Well, Google just announced a series of launches and optimization to challenge OpenAI and GPT-3: 👉 PaLM - a flexible system that can potentially carry out all sorts of text generation and editing tasks. 👉 An AI feature will soon be implemented across Google Workspace platforms, such as Docs, Slides, Sheets, and Gmail. The feature will enable auto-generated images, creative brainstorming, summarizing, and auto-generated emails. 👉 MakerSuite - to iterate on prompts, augment dataset with synthetic data, and easily tune custom models. This feature is necessary to create a consumer-friendly AI system. 👉 Vertex AI Platform - which is designed to help businesses train and deploy machine learning models. 👉 Generative AI App Builder - A platform for developers to quickly ship new experiences including bots, chat interfaces, custom search engines, digital assistants, and more. ✅ In my opinion, all these AI features will make work and businesses run even more smoothly and efficient. As a creative marketer, this innovation will help humans achieve the next level of creativity. We saw how brands like Coca Cola and Nestle are running out AI-based marketing campaigns. I'm excited to see what coming!🤩 What do you think about today's AI innovation? Let me know in the comments!👇 --- Read more of the news here: https://lnkd.in/gdXeVQJE 👉 For more marketing updates, subscribe to my newsletter❗️ https://lnkd.in/grjgYATK #PoweredbyAJMarketing #ai #marketing #google