- TOMORROW'S APE
- Posts
- The AI Gold Rush of 2024
Images in this Newsletter made with ideogram.ai
Hi, welcome to the latest issue of Tomorrow’s Ape!
It would seem Months are turning into Weeks now with new releases and announcements.
I’ve put together my highlights in the Fallen Fruit Section. These latest releases point to a huge year ahead in the building and rollout of AI.
There is no doubt we will see many applications on offer.
Stu Wilson 🤓🐵🤖
Here’s what we’ve got this week:
The latest Fallen Fruit
2024 Gold Rush
How Smart are we?
Mistral’s Mixtral: It’s a Mouthful
Some Extra Helpful AI’s worth a try
Retro Movie Rewind
FALLEN FRUIT | BITE SIZE PIECES OF AI TECH THIS WEEK |
🤖Runway the company behind the impressive RunwatML video generation AI announces their long term project to build a General World Model through their Runway Labs. What if AI’s had a understanding of the world through visuals and kept an internal world model of it’s environment? This is quite a powerful concept and the ambition for the company comes off of the back of Runway’s exception Video Generation AI. Check out this great video here, which also helps you understand the current tech behind what they do.
📦 Runway has this week given Creative Partner Program users access to Text to Speach functionality which allows you to add voiceover to Gen-2 videos.
📦 Midjourney are making their first moves off of Discord and onto their own website Midjourney Alpha. If you have created 10,000 images or more with their service you’ll get access.
📦 Remix AI are looking to do with Generative Art what Instagram did with Photos. A social media platform that allows you to easily upload your images and instead of adding filters, you can remix them. Available only on mobile IPhone and Android.
‘We’re Waiting’ by Tomorrow’s Ape
🤖 Have a look at the new Tesla Optimus Gen-2 The robot is looking far smoother than any we have seen before.
🤖 Speaking of Robots, check out this one Alter3 which “is capable of making spontaneous movement using a Large Language Model (LLM), more specifically ChatGPT4”.
🤖 Introducing Forward’s CarePod. The worlds first AI Doctors Office! The future of Health & Medicine has many possibilities now opened up by AI for monitoring and personalised analysis.
🤖 If you want to see more possibilities of what can be done with AI generated content, in this case video, look no further than this video released by Channel1. Fully AI generated News, crazy!
🤖 Check out how NeRFs are progressing. Creating 3D scenes from photos are evolving. Google Deepmind have released a paper Nuvo that helps you edit UV maps on your scene.
🤖 Apple releases a new Framework optimised for Apple’s M[x] chips as explained here by Jim Fan. It show’s an important step in Apple’s intention to integrate AI on it’s hardware. Check out this Article too.
GOLD RUSH | HOW NEXT YEAR SHAPES UP TO BE AN EXPLOSION OF AI RELEASES |
It's fair to say that given the ramping up of releases that are happening at the tail end of this year, 2024 is going to look like an AI Gold Rush. So much to be gained and the field is wide.
Open Source is punching above it's weight with releases from the likes of Mistral looking promising, but Gemini Ultra, the anticipated GPT5 or even 4.5 are likely to show their muscle early.
How much will fine tuning, compression, prompting techniques and datasets evolve in the Open Source arena and what will be the impact?
Can Open Source catch up to private companies like Open AI & Google?
What products and services will derive from the plethora of models already on offer?
Will out of the box great ideas mixed with a growing user adoption & creative utilisation see the year explode?
This year has been Seminal and Historic, next year will be all about the Building, and who doesn't love that word right now 🤓🐵🤖🚀
Some key areas to watch 👀:
🤖 Smart Appendages: Wearable AI & VR
🤖 Enterprise Adoption i.e. specific AI use cases tailored to business productivity
🤖 Medical Applications, focused more on monitoring health
🤖 Robotics including LLM integration
MISTRALS MIXTRAL | A NEW KIND OF CHATBOT FROM THE OPEN SOURCE ARENA |
This month has seen a slew of Open Source LLM’s coming out all with smaller databases and all punching above their weight when it comes to performance against larger ones. (Most of this activity has come about because of the recent AI event in New Orleans NeurlPS 2023 in which the whole industry converges to show their wears).
With various Machine Learning techniques these smaller models have gained a lot of ground which makes them a great alternative to expensive LLM’s like ChatGPT and the upcoming Google Gemini Ultra.
The French company Mistral have recently released Mixtral of Experts, which as its name suggests utilises several expert Models within its framework.
It shows another innovation in AI that would see chatbots becoming far more accurate. Rather than one huge model with a vast dataset to draw upon, Mixtral has several that are astutely fine tuned for specific subjects.
For a great, albeit simplified explanation check out this one by Matt Wolfe.
HOW SMART ARE WE REALLY? | THE CLOSING GAP IN THE HUMAN TO AI INTELLIGENCE DIFFERENTIAL |
Here is the closing gap in the Human - AI Intelligence Differential
1st it started with Winning at Chess
Then Winning at Go
…Image recognition
...Language recognition
...Predictive analytics
...Image generation
...Driving a car
...Conversation
Okay, but we are still smarter 🤓
Next…
...Emotional Intelligence
...Understanding social nuance
...Learning from limited data
...Adapting to different situations
...Moral & Ethical reasoning
...Internal monologue/narrative & planning…
…and then AI will forget it's keys and we will realise we are not as smart as we thought we were haha.
Is it possible that AI will simply ‘Mimic’ every type of Intelligence Humans possess?
One thing is clear, with current models proving that they can do a good job of mimicking many of our brains functions, we are getting closer to finding out.
YES, THIS WILL HELP | AI’S THAT ARE WORTH A TRY IF THEY MAKE LIFE EASIER |
Poe is an AI Aggregator that lets you interact with a wide range of Chatbots. It’s a great way to find the one that works for your needs. Includes the likes of ChatGPT, Claude and even DALLE3 and StabDiffusionXL Check out a full explanation here. |
Perplexity offers an alternative to using Google & Bing for search when researching. Check out this article by Forbes. |
If email is a big part of your job or business, then Superhuman might be of help. It will assist your email writing as well as provide summaries. |
Ex Machina (2014)
We’re not rewinding too far back for this one. Ex Machina is a film that reveals more about our present concerns with AI super intelligence.
It features a mysterious AI tycoon who invites a promising employee to come assess his latest inventions.
The whole film takes place in deep isolation within the mountains and it sets the stage for a challenging interplay between the humans in the story and the artificial humanoids who act and look completely human. Or do they?
Mimicry and the cold hard interior of a machine that is made in the image of its creator leads to a gut wrenching finally.
Link to IMDB here
What did you think about this weeks Newsletter?
Did we nail it? Can we do better? Want to know more? Hit reply and let us know!
If you like our content and think it might be worth a shout out, please do!
Right click the below button and share the link.