New vid : Follow & Comment "1". 1 will win a $5,000 NEO Tokyo s2. (Must like/follow/share)
The most important video I'll make this run.
We are entering the DANGEROUS part of a bull. Selling should be TOP of mind. Here's when too.
https://t.co/JZbtjvo4HT
You can now transcribe 2.5 hours of audio in 98 seconds, locally.
A new implementation called insanely-fast-whisper is blowing up on Github.
It works on works on Mac or Nvidia GPUs and uses the Whisper + Pyannote library speed up transcriptions and speaker segmentations.
Here's how you can use it:
pip install insanely-fast-whisper
insanely-fast-whisper --file-name <FILE NAME or URL> --batch-size 2 --device-id mps --hf_token <HF TOKEN>
New vid : Comment "1" below. One will win a $4500 NEO Tokyo s2. 3 win $500 USDC (Must follow n share)
This section of the gaming market (that is 90% new n undiscovered) is going to go nuclear.
Here is what it is and where I'm all in..
https://t.co/1saIdUHQbl
This site was built in 1 hr almost entirely by AI.
We prompted Relume with info about our product (an ebike).
Relume generated the sitemap, copy, and wireframes ready to be copy/pasted into webflow. All assets and icons were prompted in Midjourney.
https://t.co/KcllPqqXiC
More huge news in the 3D space, this time from Luma.
Their app now supports Gaussian splats.
Anyone with a phone can take a video and turn it into an insanely detailed 3D scene.
Crazy thing is the files are under 20mb and can be embedded anywhere 😳
Many people complain about ChatGPT's results.
But you have to write the right prompt to get a good answer.
I show you the perfect ChatGPT Prompt formula in just a few steps ↓
1. Task
Begin with an action verb, specifying the desired output.
E.g.: Generate a table with fruits to eat in summer.
2. Context
Offer just enough relevant information by addressing background, success criteria, and environment.
E.g.: I'm a content creator talking about artificial intelligence on Twitter/𝕏."
3. Exemplars
Including examples or frameworks within the prompt enhances output quality.
E.g.: (Task) Give me design ideas for my website with an object, a color, an action and a style. For example: "A blue robot sitting on a chair in a futuristic style".
4. Persona
Define who you want the AI to emulate, considering their expertise and role.
E.g.: You're a recruiter who's been working in human resources for 20 years.
5. Format
Visualize the desired output format, which can be paragraphs, bullet points, emails, etc.
E.g.: Give your answer in the form of a table with 3 columns: "Type of task", "Ease", "Degree of urgency".
6. Tone
Communicate the desired tone, style, or mood for the AI's response.
E.g. "Use a formal, professional tone to give your answer".
Impressive. MetaGPT is about to reach 10,000 stars on Github.
It's a Multi-Agent Framework that can behave as an engineer, product manager, architect, project managers.
With a single line of text it can output the entire process of a software company along with carefully orchestrated SOPs:
▸ Data structures
▸ APIs
▸ Documents
▸ User stories
▸ Competitive analysis
▸ Requirements
This is a big day. Meta is open-sourcing AudioCraft.
You can now generate incredible music and sounds with a single prompt.
It includes the most performant Generative AI Model (audio) on the market, the "Llama" of Audio.
The research framework contains the weights and code of these models:
▸ MusicGen: controllable text-to-music model.
▸ AudioGen: text-to-sound model.
▸ EnCodec: high fidelity neural audio codec.
▸ Multi Band Diffusion: An EnCodec compatible decoder using diffusion.
This is going to tremendously speed up audio research 👏
Another Midjourney workflow tip that can save you some time copy/pasting.
You can automatically append anything you want (not just parameters) after all prompts with /prefer suffix.
It might help you speed up prompting for building scenes and storyboarding.
A quick breakdown