When the studio practice becomes boring, uninteresting, or monotonous. We are left with only two options;
1) To take a leave.
or
2) To try something new.
One of the benefits of being a part-time contemporary African artist, is the freedom to paint when I want, the way I would like to (without too much external pressure).
Friday, November 24, 2023
Artist Packaging themselves for the Art World with Limited Resources: An African Artist's Journey
Friday, November 10, 2023
Grok and ChatGPT Face Off ~ AI Chat that is not politically correct
Elon Musk's introduce Ai Chat |
Elon Musk recently unveiled his latest AI chat, Grok, under his Venture XAI initiative. Grok boasts noticeably fewer restrictions compared to its competitors and is promoted as capable of answering questions with a touch of wit and a rebellious streak. The company's website humorously warns against using it if you hate humor, claiming Grok can respond to spicy questions typically rejected by other AI systems.
While the announcement doesn't clarify the exact meaning of "spicy" or "rebellious" in the context of commercial AI, concerns arise about potential risks. Most AI models are programmed to abstain from generating explicit, violent, or illegal content and minimize biases. Without such safeguards, there's a worry about AI's potential involvement in nefarious activities.
XAI's announcement mentions that Grok is built upon a language model known as Grok 1, featuring an impressive 33 billion parameters. They claim to have developed Grok in just two months, highlighting a significant advantage in real-time knowledge through the XPlatform, formerly known as Twitter, acquired by Musk for $44 billion in 2022.
Stella Biderman, an AI researcher with Uther AI, finds XAI's claims credible, suggesting that Grok likely employs retrieval-augmented generation to incorporate current information from the XPlatform into its output. Other advanced language models, like ChatGPT, also use similar techniques, relying on search engine results and other sources.
Over the past year, large language models have demonstrated remarkable capabilities, ingesting extensive text data from books and the internet. They undergo human-guided training to reduce the likelihood of generating offensive or dangerous content, enhancing their ability to provide coherent responses. However, they still remain susceptible to errors and biases. The launch of Elon Musk's new AI chat raises questions about its practical use, with skeptics viewing it as a potential hype or the next big thing after ChatGPT.