It is a common misconception that one needs immensely powerful fog computers to operate IoT services. In reality, a regular PC or Mac is fully capable of running applications that can promptly generate AI-driven content, potentially saving you from spending $20 every two weeks on services like ChatGPT Plus.
These tools are both free and user-friendly, offering a wide selection of large language models (LLMs) with specific functionalities such as text generation and code writing.
While they may not encompass the full spectrum of ChatGPT’s capabilities—like image generation or file interrogation—they excel at performing a multitude of AI tasks instantly, from composing emails and website content to simplifying complex concepts in plain language.
Moreover, these tools operate without transmitting your personal data to the cloud and come at no cost. They can even function offline without an internet connection.
Below are instructions on how to leverage two highly potent AI applications on your PC or Mac.
Jan
Jan boasts a variety of distinct AI models to choose from.
Barry Collins
Jan, a user-friendly AI software, is compatible with Windows, Mac, and Linux computers.
Upon installation, the first step is to select the desired AI model (LLM) for use. You have the flexibility to manage multiple models, most of which are available for free. (Note that while ChatGPT is listed among the available designs, utilizing it in Jan requires paid access to the OpenAI API.)
Which model should you opt for? The Trinity model is ideal for general tasks such as content creation and concept explanation. On the other hand, the Gemma model, derived from the same AI systems as Google’s Gemini AI, is geared towards experimentation.
The list of AI models in Jan indicates the amount of storage space they will occupy on your computer. While basic models may only require a few megabytes, more advanced models could necessitate 30 to 40GB. Jan will alert you if your computer lacks sufficient RAM to run certain models, although most PCs and laptops should handle the basic models without issues.
Once you have selected and downloaded your preferred model(s), you can initiate conversations with your AI by clicking on the chat bubble in the left-hand menu. The right-side window offers various settings that can be adjusted before commencing the chat.
In the Model section, you can choose which installed AI model to interact with. The Instructions area allows you to customize the responses from the AI assistant based on your preferences. For instance, if you intend to create guides for children, you can set instructions like “explain everything as if I’m an eight-year-old”.
You can now start conversing with the AI. Apps like Jan excel at tasks such as drafting business emails from bullet points, which is particularly beneficial for individuals with limited language skills or conditions like dyslexia.
For instance, by instructing the Trinity model to assist in composing business emails for a car repair business, it swiftly generated a professional five-paragraph email with all the necessary details. This exemplifies how a free AI tool can provide tangible business value.
LM Studio
LM Studio is more tailored for coders and advanced users.
Barry Collins
Similar to Jan, LM Studio allows users to select from a range of downloadable AI models, engage in AI-driven conversations, and even host a ChatGPT-like AI server for those with the technical expertise.
The interface of LM Studio is geared towards a more tech-savvy audience, making it particularly suitable for advanced users, especially programmers.
With an established community of active users on Discord, LM Studio caters to a more technically inclined audience.
Within LM Studio, users can adjust various settings in the right-hand pane while engaging in conversations, including the valuable hardware settings. These settings enable users to allocate their computer’s resources for AI processing. Moving the slider to the right can result in faster responses, while shifting it to the left can alleviate strain on the computer.
LM Studio offers access to models designed for general chat or copywriting tasks, as well as those tailored for coding purposes.
In the Playground section of LM Studio, users can simultaneously run multiple AI models to determine the most effective one. For instance, by requesting the Gemini and Llama models to provide instructions on uninstalling an app in Windows 11, I received different methods from Gemma and a more straightforward set of instructions from Llama. By leveraging multiple AI models concurrently, users can select the most suitable option for their needs!
