Grok 1 on the market

Grok 1 on the market

A few days ago (November 4, 2023) Elon Musk introduced the Grok, a new AI chatbot created by his company xAI, which is described as “entertaining and rebellious”.

From the published information, we know that Grok was developed in only a two-month training period and is intended to be not only witty but also free from censorship (whatever it means), both in “standard” and “humorous” modes.

How does it stand in comparison to the GPT-4?

The main AI engine is called Grok-1 and, according to xAI, shows an advantage over GPT-3.5 in some tests, though it currently falls short of GPT-4.

Similarly to GPT-4, Grok-1 has a ~8k context length (8,192 tokens).

The cut-off for its internal “knowledge” (the dataset based on Internet resources) is Q3 2023, so this will be also comparable to the GPT-4 turbo release (announced last days).

A unique and fundamental advantage of Grok is that it has real-time knowledge of the world via the 𝕏 platform. It will also answer spicy questions that are rejected by most other AI systems.

What is interesting, Grok has access to current content on the X platform (real-time access), allowing it to discuss current events. Having in mind that X is trying to protect “Twitter” data from being used by other companies, it might be an interesting move.

How to use Grok?

After the beta testing period, it will be made available to all Premium+ subscribers of X (formerly Twitter).

For more details on the model, see the Grok-1 model card: https://x.ai/model-card/

Sources:

https://x.ai/ https://x.ai/model-card/

Related Posts

How to evaluate meaning of colors used on the website?

How to evaluate meaning of colors used on the website?

  • Sep 13, 2023
  • TIL

I love the red color - it is probably my favorite color for sports cars (I don’t have a red sports car, but I might have one).
This is also my favorite for guitars, and I had a red one for years.

Read More
Top 10 Linux commands you must... no, this is not that kind of article

Top 10 Linux commands you must... no, this is not that kind of article

Knowing Linux and its commands is the foundation of every DevOps engineer. For devs guys that came to DevOps with a… dev background ;) it might be a bit challenging to remember all of them, as they were not using them on a daily basis - at least not as often as the Linux sysadmins. The all of them is a bit misleading, and the truth is that every person has their own set of tools in the Swiss Army Knife - sometimes it is a bit bigger, sometimes smaller - but it is always more about being able to place them in the context of your work rather than remember tons of manuals and help pages.

Read More
My AI/ML starting point, crash courses and roadmap

My AI/ML starting point, crash courses and roadmap

This is it. My starting point. I have been working with Machine Learning models for a while now, but I have never had a chance to learn the theory behind it. I have been using the models as a black box, and I have been able to get some results, or help as the engineer to build some MLOps pipelines - of course with the high level of ML knowledge required.

Read More