Nine AI Developments That Will Change the Future
Artificial intelligence (AI) is advancing at an unprecedented pace, and in the last few days, there have been nine impactful AI developments that I want to share with you. From the frankly startling Hey Gen video translation to the Epic new prompt optimizing paper, and from Apple's iax GPT to Open Interpreter Next GPT, there is a lot to cover. So, let's dive in and explore these developments in detail.
Hey Gen: Generating Lifelike Videos and Language Dubbing
You've probably already heard about Hey Gen, which can generate lifelike videos and is available as a plugin to Chat GPT. But did you know that it can also do video language dubbing? Today, I got access to their new Avatar 2.0 feature and decided to test it out with Sam Altman's testimony to the Senate. I want Spanish language speakers to tell me how it turned out. I have been researching three or four tools, including this one, to translate my videos into dozens of languages, and I can't wait to put that into place.
Open Interpreter: An Open Source Code Interpreter
Open Interpreter is an open-source code interpreter that was released five days ago. I've been using it intensively, and while it's not perfect, it has proven useful. For example, I asked it to download a YouTube video in 1440p using Pytube and clip out a specific section, and it did it in just a few seconds. This process would have taken me much longer to do manually.
Google DeepMind's Fascinating Paper on Optimized Prompts
Google DeepMind has released a fascinating paper on optimized prompts for language models. These prompts are not small optimizations, and they work with a variety of large language models. The paper says that the best prompts optimized by their method outperform human design prompts by up to eight percent on a particular math challenge and by up to 50 on big bench hard tasks. These are long-standing tasks known for their difficulty for large language models.
Google's Gemini Model: A Direct Competitor to GPT-4
Google has given a small group of companies access to an early version of Gemini, their direct competitor to OpenAI's GPT-4. According to a person who has tested it, Gemini has an advantage over GPT-4 in at least one respect. The model leverages reams of Google's proprietary data from its consumer products, in addition to public information straight from the web.
Apple's iax GPT: Designed to Boost Siri
Apple's iax GPT is designed to boost Siri, and it almost sounds like Open Interpreter, where you can automate tasks involving multiple steps. For example, telling Siri to create a gif using the last five photos you've taken and text it to a friend.
Roblox's New AI Chat Bot: Allowing Creators to Build Virtual Worlds
The online game platform Roblox is bringing in a new AI chatbot that's going to allow creators to build virtual worlds just by typing prompts. This development is going to become intuitive to the next generation, and children today are just going to expect their apps to be interactive and customizable on demand.
Smell to Text: A Narrow AI Trained in a Different Way
We now have Smell to Text, a much more narrow AI trained in a very different way to GPT models, but it matches well with expert humans on novel smells.
Protein Chat: Enabling Users to Upload Proteins and Ask Questions
Protein Chat enables users to upload proteins, ask questions, and engage in interactive conversations to gain insights.
Next GPT: A Multimodal LLM That Can Go from Any Modality to Any Modality
Next GPT is a multimodal LLM that can go from any modality to any modality. We're talking about images, audio, video, and the output being images, audio, text, or video.
As you can see, AI is advancing at an incredible pace, and these developments are just the tip of the iceberg. The world is only going to get more crazy from here, and it's up to us to navigate the future of AI.
Highlights
- Hey Gen can generate lifelike videos and do video language dubbing.
- Open Interpreter is an open-source code interpreter that can download YouTube videos and clip out specific sections.
- Google DeepMind's optimized prompts outperform human design prompts by up to 50 on big bench hard tasks.
- Gemini, Google's direct competitor to GPT-4, leverages reams of Google's proprietary data from its consumer products.
- Apple's iax GPT is designed to boost Siri and automate tasks involving multiple steps.
- Roblox's new AI chatbot allows creators to build virtual worlds just by typing prompts.
- Smell to Text is a narrow AI trained in a different way to GPT models.
- Protein Chat enables users to upload proteins, ask questions, and engage in interactive conversations to gain insights.
- Next GPT is a multimodal LLM that can go from any modality to any modality.
FAQ
Q: What is Hey Gen?
A: Hey Gen is an AI tool that can generate lifelike videos and do video language dubbing.
Q: What is Open Interpreter?
A: Open Interpreter is an open-source code interpreter that can download YouTube videos and clip out specific sections.
Q: What is Gemini?
A: Gemini is Google's direct competitor to GPT-4, which leverages reams of Google's proprietary data from its consumer products.
Q: What is iax GPT?
A: iax GPT is Apple's AI language model designed to boost Siri and automate tasks involving multiple steps.
Q: What is Roblox's new AI chatbot?
A: Roblox's new AI chatbot allows creators to build virtual worlds just by typing prompts.
Q: What is Smell to Text?
A: Smell to Text is a narrow AI trained in a different way to GPT models.
Q: What is Protein Chat?
A: Protein Chat enables users to upload proteins, ask questions, and engage in interactive conversations to gain insights.
Q: What is Next GPT?
A: Next GPT is a multimodal LLM that can go from any modality to any modality.
