Issue 43: Taming AI before catastrophe
Issue 43: Taming AI before catastrophe
6 proven practical frameworks for brainstorm; Will we have time to tame AI before a global catastrophe?; New figma plugins & more!
Hello, dear readers! 👋
In this issue, among other things:
How ChatGPT increased the productivity of business professionals by 59%
A great overview of trends in business strategies and design of beer companies
Will we have time to tame AI before a global catastrophe?
6 proven practical frameworks for brainstorm
Meta AI introduced the AI technology SAM
Speech by Yuhki Yamashita, Product Director at Figma
New Figma plugins
How the various blending modes in Photoshop work
Promts for creating logos in Midjourney
Quotes from "Things That Make Us Smart" book by Donald Norman
Enjoy reading!
📚 Book quotes
This week I suggest you another book by Donald Norman "Things That Make Us Smart: Defending Human Attributes In The Age of Machine" Review the quotes and decide whether to read or not:
We still suffer from the mind-set of the Chicago World’s Fair: “Science Finds, Industry Applies, Man Conforms.
It is easy to devise numerous possible scenarios of future developments, each one, on the face of it, equally likely.
The personal point of view: Artifacts change the task. The system point of view: The person + artifact is smarter than either alone.
In a proper system, the process of exploration will let us discover the question as well as the answer.
Humans need a meaningful, accessible representation: sounds, sights, touch, organized in meaningful, interpretable ways. The result, however, is that we are ever more dependent upon the design of our devices to make the information visible and to make the artifact usable.
There are two major tasks for the user of an information display: 1. Finding the relevant information; 2. Computing the desired conclusion.
Grudin’s law: When those who benefit are not those who do the work, then the technology is likely to fail or, at least, be subverted.
Appropriateness principle: The representation used by the artifact should provide exactly the information acceptable to the task: neither more nor less.
As a result, the two together—the powers of the machine and the powers of the person—complement one other, leading to the possibility that the combination will be more fruitful and powerful than either alone. If we design things properly, that is.
How have we increased memory, thought, and reasoning? By the invention of external aids: It is things that make us smart.
The problem with stories is that they are too flexible, too subjective: Almost any point of view or conclusion can be buttressed by an appropriate story. In the end, we often need both—the hard, formal process of logic and the soft, subjective impressions from everyday experience.
Try to hold onto too many thoughts at once, and they interfere with each other. The psychological evidence is that only about five things can be kept active in consciousness at a time. Five.
Technology should be our friend in the creation of a better life; it should complement human abilities, aid those activities for which we are poorly suited, and enhance and help develop those for which we are ideally suited.
Technologies are not neutral. They affect the course of society, aiding some actions, impeding others, independent of the morality or necessity of those actions. Technology also has its side effects, both physical and mental. Technology can aid as much as it can detract. It really is up to us, both as individuals and as a society, to decide which course we shall take.
🗞 News and articles
Humanity vs Artificial Intelligence. Will we have time to tame AI before a global catastrophe?
Big article is about the pace of AI development, the potential threats it can pose to humanity, how to deal with them and why people underestimate the speed of technological progress.
Conceptually, there are two types of AI: a highly specialized "ANI", which is taught by people, and a universal "AGI", which will be able to learn independently and perform any task that a person can perform. The development of AGI could potentially lead to a global catastrophe if such an AI does not take into account human values. Therefore, the concept of "AI alignment" was invented, which implies the creation of AI taking into account the values of humanity. However, Vastrik points out the difficulties of implementing this concept and explains why AI will find workarounds and why human ethics cannot be reduced to simple rules
The Artificial Intelligence Revolution: Part 1 - Wait But Why
2023 Beer & Beverage Branding Trends Review
A large and detailed overview of trends in branding and business strategies of beer companies in the USA.
Business trends:
Breweries often have to rebrand not only to refresh the design, but also to stay relevant
When choosing beer, buyers are guided by brand credibility, not stylish packaging
Manufacturers often create packages that are designed as sets for specific events. For example, for a party or birthday
The young audience is not interested in traditional beer and generally refuses alcohol more often. At the same time, research data vary, and some of them claim that generation Z drinks as much alcohol as any other
Manufacturers are experimenting with the amount of beer in the package. Now you can more often find packages with 12, 8 and 4 bottles or cans
For the first time in history, women drink more than men, while the proportion of black and Hispanic women who regularly drink alcohol has increased by 84% since 2004. This may affect the strategy of brands that used to be focused on white male millennials
Companies are more likely to focus on creating sub-brands that are promoted and positioned independently
Visual trends:
Mascots
Abstract figures
Illustrations with the ingredients of the drink for the convenience of taste selection
Nostalgic motifs with references to various manifestations of pop culture
Gender-neutral design
2023 Beer & Beverage Branding Trends Review - CODO Design
ChatGPT Lifts Business Professionals’ Productivity and Improves Work Quality
NNGroup was told about a study by scientists from the Massachusetts Institute of Technology, who found out that people using ChatGPT to write business documents work faster and better than those who do not use neural networks.
The study involved 444 business professionals from various fields, including marketers, data analysts and HR specialists. Each respondent was asked to write two business documents in their field with and without AI, with 70% of participants using AI for the first time. Participants using ChatGPT prepared their result in an average of 17 minutes, while specialists who wrote their document without AI support spent 27 minutes on it. The result was a 59% increase in productivity, while the quality of documents written with AI turned out to be higher.
The use of AI has also changed the distribution of working hours. Without the help of AI, participants spent about 25% of their time brainstorming, 50% writing a draft, and 25% editing a draft. When using AI, the time for creating drafts was halved, but the time for "polishing", on the contrary, doubled.
ChatGPT Lifts Business Professionals’ Productivity and Improves Work Quality
⚡️ Briefly
Unlocking ideas: 6 frameworks for design brainstorming. Readymag talked about such proven practical frameworks for brainstorm as the Walt Disney method, the 6-3-5 technique, mashup, silent storm, dueling framework and step-by-step presentation.
Unlocking ideas: 6 frameworks for design brainstorming
New technologies
The non-profit organization Future of Life has published an open letter in which it called for suspending the training of AI systems more powerful than GPT-4, as they can carry serious risks for humanity. They propose suspending training for six months so that AI labs and independent experts can use this pause to jointly develop and implement common security protocols. The initiative group included about a thousand scientists, entrepreneurs, writers and so on, including such famous personalities as Elon Musk and Steve Wozniak.
Pause Giant AI Experiments: An Open Letter - Future of Life Institute
Image to Text for Midjourney AI. Now Midjourney can generate promts based on the reference image. The new function can be called using the /describe command, after which the AI will provide 4 options for text prompts to create a similar image. The authors of the note conducted their experiment and showed the results.
Image to Text for Midjourney AI - AI Demos
Meta AI introduced the AI technology SAM (Segment Anything Model), which allows you to very accurately recognize individual parts of objects and their sizes. For example, she can identify the handle on the door, the book on the table and much more. The new technology could potentially be used in robotics and healthcare.
There is an interactive demo on the site where you can test the technology on various photos.
🧘 Inspiration
Branding
Cohere. The original identity of an AI platform for business with a pleasant modern look, a restrained palette and 3D illustrations. As the main visual metaphor, the authors used the structure of living cells and Voronoi diagrams, on the basis of which patterns, animations, a layout system and even a custom variable font were created. It turned out to be a very fresh and extraordinary style.
Laconic identity of the Panzon diner with a hand-drawn logo and a minimalistic mascot.
Askly. A pleasant identity of chat developers for online stores with an animated logo, an original style of illustrations and a pleasant gentle range.
🌁 Posters
Colorful and diverse posters by Elvira Akatieva in a brutal avant-garde style.
Elvira Akateva - typo/graphic posters
🖨 Printing
Studio Vernacular. A book about cyclical trends in European design with a bold and modern layout.
Keep reading with a 7-day free trial
Subscribe to bezier.design to keep reading this post and get 7 days of free access to the full post archives.