Newsletter issue #2 - Collective generative AI predictions for 2024
TL;DR
In this second newsletter we’re excited to share the Time Under Tension Collective with you. Over the past few months we’ve been collaborating with the brightest generative AI brains in the market. These fine folk are now part of the Time Under Tension Collective, each bringing their unique perspective to the generative AI boom. We’re delighted to introduce them here, and share their insightful generative AI predictions for 2024.
2024 Generative AI Predictions
Ben | Beyond Conversational | “We all know how to chat. It’s the ultimate interface with human-to-machine instruction and dialogue. Or is it? Customer experience and interface design will have a much-needed shakeup as real-time personalisation can truly be achieved. Will the hamburger menu be no longer? Will everyone’s homepage be their page? I’m looking forward to breakthrough interfaces that are multi-modal ingesting type, speech, vision and more. With the much needed ability to pickup where we left off we’ll trigger an exciting new chapter in interaction design”
Jackson | Big strides towards AGI | “OpenAI's second DevDay comes with an announcement regarding a big leap forward in AGI. (Q* algorithms make LLMs excellent at solving math and logic problems) Rather than LLMs being able to "predict" the answer using the most likely next token, LLMs can now use reasoning capabilities to deduce the answer with a high amount of accuracy. This would mean that we could truly leverage the power of generative AI to help us solve problems in physics, math, medicine, and biology”
Tim O | AI personality tests| “In 2024, I predict the creation of a “Myers-Briggs” test for AI’s, such that people can choose an AI assistant that is the best fit for their personality. Very soon, every LLM will be insanely clever to the point where the established technical benchmarks are meaningless (“95% on the MMLU!”, “169 billion parameters!”, “2001k context window!”).
“My prediction: a new benchmark will be created, one that measures not how clever your AI is, but its personality and behaviour. If I’m to have an AI Assistant as my travel guide, do I want it to be upbeat (ChatGPT), empathetic (Pi) or sassy (Grok)? How about for my AI math tutor? Sure, you can always instruct an LLM to ‘act like’ someone else, but wouldn’t you prefer your AI assistant to be its true self?”
A slide from Time Under Tension Inform session covers some of the different LLM’s available today
Tim F | AI Adoption Paralysis | “In 2024, despite the growing availability of advanced generative AI, many businesses will struggle to effectively use these technologies.
Overwhelmed by the vast choice and fast-paced evolution of AI, companies will face indecision in adopting AI.
Key factors contributing to this stagnation include:
Choice Overload: The plethora of AI tools creates decision paralysis, as businesses find it hard to pick the most appropriate solutions.
Fear of Obsolescence: Rapid AI advancements lead to concerns about investing in soon-to-be outdated technologies.
Integration Challenges: Difficulties in blending new AI with existing systems deter especially smaller businesses.
Skill Gap: A lack of skilled personnel aggravates the issue, with firms struggling to utilise complex AI technologies.
Economic Caution: In a climate of economic uncertainty, businesses may prioritise short-term survival strategies over investing in AI, perceived as a risky, long-term commitment.
This trend suggests a need for a more informed approach to AI adoption, focusing on education, gradual integration, and strategic partnerships. Without this, many businesses risk missing out on AI's transformative benefits due to indecision and poor planning”
Sequoia's Generative AI Market Map highlights the fast changing generative AI landscape
Hayley | Generative AI drives healthier lives | “My predictions for 2024 are around the shift in how we can (hopefully) up-level our health and wellness. There will be greater interconnectivity between AI, apps, and wearable tech, giving us proactive tools and tips to level up our sleep, diet, food and movement to achieve our health and wellness goals. Imagine your Oura ring, Noom, Outlook and gym tracker all talked to one another. Based on your readiness score, you were able to easily alter your macros for the day, adjust the type and difficulty of your workout and switch around your meetings and tasks to match your energy and alertness level”
Simon | Multimodal marketing | “2024 is the year AI-generated content puts on its serious pants. With Open AI's recent releases, Google Gemini on the way, and who know what next, there will be some mind-boggling opportunities for savvy marketers and content creators. The big one will be multimodal marketing and content - being able to strategise, conceptualise and create with a single prompt, bringing together knowledge bases, latest data, text, images, video, speech and sound to produce interactive, engaging and immersive content. I'm also excited about storytelling with image generators. And if all my predictions are wrong, I can always build an army of AI TikTok Influencers”
Google Gemini showed the power of multimodal AI’s (and the backlash when you fudge a demo 😜)
Fiona | Rise of the AI Designer | “2024 is going to bring about the rise of the AI designer. A brave new brigade who will not only look for ways these tools can augment their practice and expand their capabilities (and their teams) but also contribute to a bigger conversation about the role designers can play in building thoughtful, useful and ethical AI empowered products.”
Brad | Real world applications | “We are going to move away from cool experiments to real world applications. When the iPhone was launched there was a handful of “fun” apps that showcased its features. Remember the beer app or the lightsaber app. Before long we saw just how useful the accelerometer could be. AI will be ubiquitous. We are already seeing AI integration with a handful of apps and even in the UI in Windows. Many users understand the power of ChatGPT but it will reveal its true potential when it is integrated into the products we use every day. This will lead to a decline in the use of the ChatGPT app itself”
Microsoft Copilot will soon see generative AI appear in all common productivity tools
Josh | AI-driven platforms | “In 2024, Generative AI could revolutionise business models by enabling the creation of entirely new industries and services.
Imagine AI-driven platforms that can autonomously run businesses, from market research to product development, marketing, sales, and customer service. These AI systems could potentially start and manage their own companies, entering markets with high-speed efficiency and minimal human oversight.
We might see the rise of 'AI entrepreneurs' that can identify gaps in the market, launch products, and adapt strategies at a pace and scale beyond human capability. This could dramatically reshape the global economy, blurring the lines between human and AI-driven entrepreneurship”
Jason | Metaverse reborn | “What do you get if you combine powerful VR/AR tech (like the recently released Oculus Quest 3, and soon to be released Apple Vision Pro) with generative AI text-to-3D capabilities? A powerful combination, unlike any seen before!
Blockade Labs shows how easily text and sketches can be rendered into 3d world
The exciting promise of Metaverse (think Ready Player One!) turned out to be far from the reality we ended up with (like the failed Meta Horizons Worlds). The reasons for this range from hardware not being quite up to scratch to the high cost of creating 3D environments - meaning actual (achievable) use cases are few and far between.
But, generative AI could force a total reinvention of how we think of, work and play in the Metaverse and might just be the catalyst needed to make it take off in 2024.
Tools like Blockade Labs showcase how quick and easy text-to-3D (or sketch-to-3D) can be to create 3D environments in seconds, with little or even no 3D experience. For 3D artists, Unity announcing a suite of generative AI tools means that we can expect to see an explosion of generative AI enabled VR/AR worlds emerging next year.
Imagine instructing Apple Vision Pro to “take us back to Bali holiday 2019” to be immersed back in a lifelike memory from past, or “take me to the meeting room” to be immersed in a meeting room with your colleagues, all with a few simple voice commands, with generative AI constructing the immersive 3D worlds on the fly. Exciting stuff, made easy with generative AI”
A handful of Gen AI news
Here are five of the most interesting things we have seen and read in the last week;
1. Channel1 is a new personalised global news network powered by generative AI.
2. Outfit Anyone is an “ultra-high quality virtual try-on for Any Clothing and Any Person”
3. In our last newsletter we shared our AI Agent, today we share Relevance AI a no-code platform setting out to make it easy to “Build your AI workforce. Multiple AI agents working across multiple tools.”, a local Australian startup who just raised $15 million.
4. Microsoft launched Copilot Studio, a low-code interface which makes it easy to build your own custom Copilots and GPTs. It was launched at Microsoft Ignite, which was filled with generative AI goodness.
5. StoryMe, a creative video agency just launched the beautiful Latent Land “a homage to both human creativity and the capabilities of artificial intelligence” - enjoy!
That’s a wrap for Issue #2. Drop us a line to say hello and let us know what you think of our predictions.
