Hello, dear readers! đ
In this issue, among other things:
What is the Jevons Paradox
Personal experience of launching, developing and then selling a Chatra startup
Evaluation of four possible scenarios for the use of AI in user research
How ChatGPT influenced programming, SMM, design and writing freelancing
U.S. Army General on the Board of Directors of OpenAI
The new most powerful open source programming model from Nvidia
A new popular platform for designers and artists with an AI filter
âŚand much more!
Enjoy reading!
đ News and articles
Lessons Learned After Selling My Startup
Yaakov Karda spoke about his experience of launching, developing and subsequently selling the Chatra startup, and most importantly, shared valuable conclusions.
He talked about competition, market entry strategies, attracting and retaining customers, evaluating the company, retaining employees after the sale and other subtleties.
User Research with Humans vs. AI
Jacob Nielsen analyzed all possible scenarios for using AI in conducting user research. He highlighted the main roles of people in each study and told which of them could be occupied by AI and which could not.
Who is AI Replacing? The Impact of Generative AI on Online Freelancing Platforms
A group of scientists conducted a study to find out the impact of generative AI on the freelance order market. To do this, they processed a large amount of data from the world's leading freelance platform, as well as data from Google Trends. In total, two million freelance publications with orders from 61 countries published from July 2021 to July 2023 were analyzed.
The results showed a decrease in the number of orders in popular areas:
-21 % of orders for writing code and text
-13 % SMM services
-17 % graphic design and 3D
Jevons paradox is not surprising
Seth Godin recalled the Jevons paradox, according to which, as the efficiency of a resource increases, its consumption increases, not decreases.
For example, with the advent of washing machines, we raise the bar for cleanliness and wash more. In this logic, with the advent of AI coding, the demand for a good year will increase and even more programmers will be needed.
New technologies
OpenAI appointed Paul Nakasone, a retired U.S. Army general, to the board of directors. Until recently, he was the director of the US National Security Agency (NSA), as well as the head of the US Cyber Command, whose main tasks are conducting cyberwar operations and protecting US military computer networks.
Nakasone was also one of those who fiercely defended the extension and tightening of the U.S. Mass Surveillance Act (FISA). Many consider the act to be quite controversial, but it was still extended in April 2024.
According to the company, Paul Nakasone will be responsible for cybersecurity at OpenAI, but a number of commentators, including Edward Snowden, believe that the purpose of such an appointment lies in the military use of AI. Until January 10, 2024, the company completely prohibited the use of its services for the development of weapons and military purposes, but in the new version of the rules, only the use of "harm" is prohibited. Thus, a loophole opens up for the use of technology by the US military departments.
It is possible that Ilya Sutskever and Jan Leike, who were involved in AI security and were concerned about the threats of using AGI against humans, left the company because of plans to appoint Nakasone to the board of directors. Earlier, I mentioned about the situation inside the company and about Ilya Sutskever's resignition due to security problems in OpenAI.
Just before the publication of this issue, the news appeared that Ilya Sutskever, together with his colleagues, is launching the company Safe Superintelligence Inc. According to them, they will create a secure superintelligence (SSI) without the distraction of commerce. That is, to do what OpenAI originally planned.
Multiple Goals of Safe Superintelligence Inc:
To assemble a highly qualified team of engineers and researchers and fully direct their efforts to the development of SSI
Develop SSI as quickly as possible. At the same time, security will have to outpace the development of SSI
To ensure that safety and progress are protected from commercial pressure
Nemotron 4 340B. The new most powerful open source neural network for programming from Nvidia. It is positioned as a competitor to GPT-4o. Knows more than 50 languages, 40 programming languages and can be used for commercial purposes. It has already surpassed Mixtral 8Ă22B, Claude sonnet and Llama3 70B in tests. It will take more than a terabyte of hard disk space to run locally.
Online access should appear later on the Nvidia website.
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.