AI Rewind: Making PowerPoint Presentations with AI

This week’s most Interesting AI news

December 23, 2023 • Estimated Reading Time: 6 minutes

Roll Up of This Week’s AI Posts

Here’s my second week of AI Rewind, a curated list of posts and links to things I thought were interesting this week. Enjoy!

While the dream of "one-shot" AI presentations isn't yet a reality, I've found a sweet spot. By integrating AI tools into my workflow, I've cut down presentation creation time by a third.This isn't just about speed; it's about quality. AI helps design visually striking, impactful presentations that truly resonate with audiences. Here's how I use ChatGPT, Gamma, Beautiful.ai, and PowerPoint to create presentationsI wrote a more comprehensive guide here if you are interested. https://lnkd.in/g_4YXKa6

Quick Favor?

If you like the content and read this email in the SubStack App, you can click the heart if you like the post. Or, if you read the post on LinkedIn and like it, can you link or reshare it? That’s the best way to support this mailing list.

One of the most powerful features of ChatGPT is its understanding of sentiment. I often use it to rewrite an email when I am too tired to exercise the tact I'd like. A lot of times it allows me to respond without the cognitive load of trying to say politely that this communication stinks or I disagree.

More so than anything else, when I talk to my friends about AI, they almost always share one of two sentiments, “I am not sure we can trust AI.” or “How soon before AI takes my job?” I tell them that despite its impressive capabilities, AI still has a long way to go before it can match the cognitive abilities of a three-year-old child or even a dog.

Last week, thousands of artificial intelligence researchers, engineers, founders, and investors flew to the Big Easy for the Conference on Neural Information Processing Systems— @nipsconference.There was a lot of buzz around a new open source LLM with an architecture that differed slightly from the popular transformer models like OpenAI's GPT-4. This could be a big deal as smaller, capable models can work on more commonly available hardware. Mamba is a new state space model architecture showing promising performance on information-dense data such as language modeling, where previous subquadratic models fall short of Transformers.

Artificial Intelligence is upending the world, so buckle up for a seismic shift in the job market. 💼 Goldman Sachs forecasts AI replacing around 300 million full-time jobs by 2030. But don't let that worry you. 💡 The World Economic Forum paints a brighter picture, promising 97 million new jobs from the AI revolution by 2025. 64% of managers in an UpWork survey acknowledge this, ready to expand their teams with AI-savvy professionals. We are at the beginning of a phenomenal opportunity—a chance to acquire new skills and boost our relevance in the AI-driven workforce.

Artificial Intelligence and how it may affect our capacity for thinkingAn article about artificial intelligence in which Dr. Nick Wagget focuses on the human side of the relationship and how this might be affected as the power of AI develops. He offers some brief thoughts on how our thinking about AI might be lost and gained in the future.

When ChatGPT came on the scene, it was nothing short of amazing. However, over time, I think the quality of the answers I get has become less helpful in creating content and researching. I also see more errors and network timeouts due to its popularity. However, there are many capable alternatives to ChatGPT. Here are five of them and why I like them.

Solar rises ☀️ Upstage just released Solar a 10B an open LLM outperforming other LLMs up to 30B parameters, including Mistral 7B. 🤯 Solar achieves an MMLU score of 65.48, which is only 4 points lower than Meta Llama 2 while being 7x smaller.

This is an excellent roll-up of Reuven Cohen's predictions of GenAI for 2024. I like these three points:OpenAI releases an open source foundational MoE model.I hope this one happens as it cements that open source development is the best way to develop and keep AI on the right track. Apple releases iOS LLM (Sub 1B parameter)Without a Jony Ives or a Steve Jobs, I hope that Apple has a wunderkind in AI to drive the smartphone in the AI industry. I think they will benefit from leveraging open source development like they did with FreeBSD for Mac OS. Major healthcare breakthroughs enabled by AI.This is the most important in the long term. Cures for diseases and more personalized medicine not dictated by big pharma are essential. AI should make research and development more accessible and not require nearly the capital to solve big problems, driving down the R&D costs and the significant premiums commanded for many drugs. Also, cures rather than treatments will become more common as the healthcare market democratizes.

Reply

or to participate.