Newsletter: July 2023

In this newsletter, we pick up on what has happened during the summer. For many, July means summer vacation, so it is for us at Viden.AI. This also means there have been no new articles for some time, but we have several articles ready for August.

For now, however, as I said, we will start with an update on what has happened over the summer.

The big tech companies are in a race to develop new advanced versions that can take over the market. Google, in particular, is working hard to develop the next language model that can take on OpenAI. They have just launched Google Bard in Denmark, and you can test it at https://bard.google.com. A Google Account is required to use Bard. Unlike ChatGPT, Google Bard has access to the entire internet.

Llama 2
Meta has also launched their new language model, Llama 2, in collaboration with Microsoft. The Verge has called the launch of Llama 2 the most important release since ChatGPT came out last November, as it is an open-source model. This makes it possible to integrate Llama into our products, and it will be interesting to see how learning resources can benefit from the new opportunities. However, it should be mentioned that while Meta would like us all to think it's an open-source model, there are several restrictions for larger companies.

You can test Llama 2 yourself via Perplexity.ai. However, Llama 2 is not yet on par with ChatGPT.

Claude 2
We also want to focus on Claude 2, which is better than ChatGPT. Claude was developed by Anthropic and started by former employees of OpenAI, with a focus on creating a safe and ethically sound language model.

One of the great advantages is that open-source software can work with large amounts of text. You can upload large files, such as books in PDF versions, and query the content, and Claude 2 can handle up to 75,000 words in one prompt.

Claude 2 can also identify text emotions and imitate human empathy, compassion, and sensitivity in conversations. In this way, it can react and adjust the tone of the answers to seem more human.

If you want to try out Claude 2, you need to create a free account with Anthropic.

Claude

Talk to Claude, an AI assistant from Anthropic

Upcoming articles on Viden.ai

AI Regulation

We are just finalizing an article on AI Regulation and what it will potentially mean for the education and training sectors. It's a complicated area, and it's a bit difficult for non-lawyers to figure out all the paragraphs. Fortunately, we have received help from Professor Dr. Jur. Henrik Palmer Olsen to ensure correct content.

Ethical aspects of chatbots in education - the uncanny valley of the mind

We have delved into the ethical challenges that arise when artificial intelligence simulates emotions and how this affects the user. There will be a time when we don't know if it's a human or an artificial intelligence we're writing or talking to.

In education, we should avoid giving artificial intelligence human qualities, and we should start with how we refer to the systems. For example, we must avoid phrases such as "artificial intelligence teaches us something". Instead, we should write, "the system is designed to...". At the same time, it is important to clarify that AI does not listen and understand but receives an input processed from a dataset that produces an output. They must understand that technology has no magic or humanity – just a programmed language model.

New Articles

Since the last newsletter, we have published three new articles, each of which has required quite a lot of preliminary studies and work before they were completed. We have prioritized these articles over writing about the latest AI technologies because many others do.

AI literacy: from SAMR to TPACK to CoACT

In this article, we look at what to consider if you incorporate artificial intelligence into teaching and how it will affect teaching. We look at the concept of AI literacy and examine the inclusion of artificial intelligence in education based on the SAMR and TPACK models. We will also look

Viden.AIClaus Scheuer-Larsen

Recommendations on using AI in education institutions

We hear from many educational institutions that are having difficulty navigating the handling of artificial intelligence and are trying to make local regulations. The big problem with regulating the use of artificial intelligence is that the technology is constantly evolving. As a consequence, the rules must be constantly updated, and

Viden.AIMalte von Sehested

Exam papers and tests: Can we detect texts generated by ChatGPT?

Every time we’re out giving presentations, we’re always asked the same question: “How do we find out if students have turned in an assignment written by ChatGPT?” In particular, it is about what we do for this summer’s written exams. Briefly about the challenge Currently, artificial intelligence is being built

Viden.AIPer Størup Lauridsen