INKISH.NEWS -

During the past weeks, I have enjoyed listening to a fantastic Danish podcast series about the physicist Niels Bohr. In 1922 he received the Nobel Prize in Physics for being essentially involved in the foundation of atomic structure and quantum theory. Topics, for most, are challenging to understand and even more difficult to comprehend. Nevertheless, his work and research are fundamental to our understanding of the science that has given us transistors, leading to computer chips and so many other things that I was stunned, from biology, and DNA, to the development of chemistry and, well, essentially crucial to how we live today!

Today's topic has nothing to do with Bohr specifically. Still, the time he lived gave us several ground-breaking innovations and understandings that today are taken for granted. Artificial Intelligence, or AI, has been around for years, and we have seen it also in the graphic arts industry.

I even believe that some have considered AI a buzzword that didn't mean anything to them or more of a marketing gimmick. The word or the understanding of artificial intelligence goes back to Frankenstein, written in 1818, or to the Czech author Karel Capek, who also wrote about artificial intelligence. His brother is recognized for being the inventor of the word robot, which origins from the Czech word robota, which simply means labor or hard work!

Therefore, the idea of having self-thinking machines is not new—Artificial Intelligence, also known as Machine Learning, is maybe a better expression. AI is an algorithm able to learn from the input given. Many may remember the IBM Watson machine, which was seen as one of the first useful AI machines and gave us all an idea about the possibilities of AI. A funny anecdote is that the name Watson possibly derives from the first CEO of IBM, Thomas J. Watson. Watson was born in 1874 and developed IBM into an international giant from 1914 till his death in 1956.

However, Watson - the computer - could learn medicine, chess, food recipes, legal - practically anything rule-based. Watson got a lot of media attention and gave the general public an idea about the power of computers, and power is the keyword.

The more complex the Machine Learning algorithm is, the more ' horsepower' it needs.

In 1929 Gordon Moore was born. Together with Robert Noyce, he founded Intel in 1968, and that closes the loop. Without Niels Bohr, no transistors, no transistors, no Intel - and without developing faster chips, no IBM and no Moore's law have evolved. Moore's law has pretty much since 1968 been true to its prediction of doubling the processor speed every 18th month - now for fifty years. Amazing, right?

With AI and Machine Learning, speed is essential. Though we don't have the processing power for AI in our computers and smart devices, we have the internet, which can deliver queries and answers to significant servers capable of processing and delivering stunning results.

As mentioned, AI is seen everywhere, and what started as simple tools, the past weeks have shown a significant startup - and now potentially disrupting the world we know. The company is called OpenAI, and its founder and CEO, Sam Altmann, is a new rock star in computer heaven. The company was founded in 2015 by Sam Altmann, Peter Thiel, Reid Hoffman, and Elon Musk, who pledged $1 billion to the project. In 2018 Musk withdrew from OpenAI. In 2019 Microsoft and Matthew Brown Companies invested in OpenAI an additional billion dollars leaving OpenAI now with a value of 20 billion dollars (430 employees).

According to OpenAI's website, the company is organized and governed as copied here: "OpenAI is governed by the board of the OpenAI nonprofit, comprised of OpenAI LP employees Greg Brockman (Chairman & President), Ilya Sutskever (Chief Scientist), and Sam Altman (CEO), and non-employees Adam D'Angelo, Reid Hoffman, Will Hurd, Tasha McCauley, Helen Toner, and Shivon Zilis."

OpenAI has developed amazing technology and gained enormous interest from companies, governments, and regular users. I have checked three services Codex, DALL·E, and ChatGPT. All are genuinely amazing, and when I write that AI will disrupt the world, you can check the three services. You will most likely sit back, speculating whether this is good or bad - and, regardless, embrace AI and see how you potentially can use it in your business.

Codex lets you program computers in almost layman's terms, enabling the development of computer programs, apps, and websites.

DALL·E generates art - and though I believe there is more to art than the visuals, this is what it does. You write Rhino, Sunset, and Motorbike, and you'll get the following image (this time).

With ChatGPT, you can ask almost anything and get very complex, and for the most get accurate information back, written in easy-to-understand English.

OpenAI has taken AI to the next level, and you will see, or have already seen, endless examples of how ChatGPT can answer complex questions. As the model learns, it will improve, and though the OpenAI disclaimer is clear about the limitations of their services, people have already started using all their services. The quality of the conversation is so good that an artificial 'person' spent 'his' Christmas on Twitter discussing politics without ANYBODY realizing they were chatting with a robot.

OpenAI's CEO Sam Altmann said in a video that he believes there will be a limited number of very comprehensive and massive AI providers as the technology requires a lot of computer power, but also monitoring, further development, etc.

He also said that the strong AIs would be industry-based solutions that teach the AI the industry details - but this is already happening using the vast AI providers.

Solutions like Descript and Runway use OpenAI in their products. The online video editor Runway enables users to add background details that don't exist but that are generated on demand. It also allows you to remove elements from an entire film, and the AI will recognize the element throughout it.

What we see now is only the tip of the iceberg. AI will 100% give humanity a lot of great things - though some will see their work disappear. I couldn't help smiling as The Economic Times, July 5th, 2022, published an article where the author Satyam Sharma, believes AI can't replace the following jobs: "Psychologists, caregivers, most engineers, human resource managers, marketing strategists, and lawyers are some roles that cannot be replaced by AI anytime in the near future.”

I wouldn't bet on that, but regardless, AI will give us new work, new challenges, and new solutions.
In China, the government has recently passed a law that requires artificially made content to be disclaimed. The problem is most likely not the disclaimed content but the content that people with bad intentions may create artificially - they will not disclaim. How we will differentiate content in the future from AI is a big question, so why not ask ChatGPT?

I asked ChatGPT: "How will humans differentiate content created by humans or by AI in the future?"
And here is the answer: "There are a few ways that humans might differentiate content created by humans or by AI in the future:

By the style or tone of the content: AI-generated content is often more formal and less personal than content written by humans, so it may be possible to identify AI-generated content by its style or tone.
By the way the content is presented: AI-generated content is often presented in a more structured or formulaic way than content written by humans, so it may be possible to identify AI-generated content by its presentation.
By the use of metadata or other indicators: In some cases, content creators may include information about whether the content was created by a human or by an AI in the metadata associated with the content.
By the content itself: In some cases, the content itself may be a clue as to whether it was created by a human or by an AI. For example, content that is highly technical or contains factual errors may be more likely to have been generated by an AI.

It's important to note that as AI technology continues to improve, it may become increasingly difficult to differentiate content created by humans or by AI."

AI is here to stay, so the good question is how we can use it for the better. Interesting times!

Jan 04, 2023

Author: Morten B. Reitoft