The Sunday Brew #60
In this brew - Green Future Index 2023 Ranking in a Picture | Moravec's Paradox & Explore-Exploit Trade-off | 23andMe Data Breach, OpenAI in License Deal with Publishers & Tesla Major Recall in China
Welcome to The Sunday Brew, weekly 1-2-3 newsletter by The Percolator. Every Sunday we drop in your inbox 1 story in a picture, 2 concepts, ideas or frameworks to expand your horizons and 3 news from the week, to keep you updated.
If you are not a paid subscriber, here is what you missed last week:
ONE STORY IN A PICTURE
TWO IDEAS, FRAMEWORKS OR CONCEPTS
This week we bring to you two concepts - Moravec's Paradox & Explore-Exploit Trade-off
Moravec's Paradox
Moravec's Paradox is the observation that, contrary to the common assumption that complex tasks require high-level cognitive skills and are difficult for machines, it is often the simple sensorimotor skills that are the most challenging for artificial intelligence.
Tasks that are easy for humans, such as walking, recognizing faces, or picking up objects, tend to be difficult for AI systems, while tasks that are difficult for humans, like complex mathematical computations, are often easier for AI.
This paradox was formulated by Hans Moravec, a pioneering researcher in the field of robotics, and it highlights the differences in the computational strengths of humans and machines. Moravec argued that evolution has optimized human brains for tasks that were crucial for survival, such as navigating the physical world and interacting socially, rather than for abstract reasoning or formal logic.
Understanding Moravec's Paradox is important in the context of AI development, as it suggests that what may seem simple for humans can be quite challenging for machines, and vice versa.
๐
Explore-Exploit Trade-off
The explore-exploit trade-off is a concept often encountered in decision-making and optimization problems, particularly in the field of reinforcement learning, machine learning, and artificial intelligence.
The trade-off refers to the dilemma of choosing between two competing strategies:
Explore: Gather more information or try new options to discover potentially better solutions. This involves taking actions that are not currently considered optimal or well-known.
Exploit: Utilize the current knowledge or best-known strategies to maximize immediate gains. This involves sticking to what is already known to be effective.
The challenge is finding the right balance between exploration and exploitation. If you focus too much on exploration, you may miss out on exploiting the best-known options. On the other hand, if you only exploit known strategies, you might miss out on potentially better solutions that could be discovered through exploration.
This trade-off is present in various real-world scenarios, such as in business decisions, game playing, and algorithmic optimization. Striking the right balance depends on the specific context and the goals of the decision-maker or the learning system.
THREE NEWS FROM THE WEEK
23andMe Shifts Blame for Data Breach to Users in midst of Class Action Lawsuits
In recent months, 23andMe, a major player in genetic testing, has been probing the details of a data breach that occurred in October, affecting millions of its users. In response to a string of class action lawsuits from breach victims, the company is allegedly shifting the responsibility back to the users, suggesting they should have exercised greater caution when reusing their login credentials.
Initially, the hackers gained entry to approximately 14,000 accounts by exploiting previously compromised login credentials. Subsequently, they leveraged a feature within 23andMe to extend their access to nearly half of the company's user base, equivalent to around 7 million accounts. Read More ยปยปยป
โ
OpenAI in talks with Publishers for Content Licensing following NYT Lawsuit
Following the lawsuit filed by The New York Times against OpenAI and Microsoft for using its articles without permission, OpenAI turned its attention to licensing content directly from publishers. This marks a significant shift in their approach to data acquisition for training their AI models.
Previously, OpenAI had been using a method called "web scraping" to gather data from the internet, which involved automatically copying and downloading content from websites without their consent. This practice, while common in the field of AI research, has raised concerns about copyright infringement and fair use of intellectual property.
OpenAI has already secured licensing deals with several major publishers, including Axel Springer SE (parent company of Politico) and The Associated Press. They are reportedly in talks with dozens of other publishers, indicating a strong commitment to their new approach. Read More ยปยปยป
โ
Tesla Recalls 1.6 Million Cars in China to over Autopilot and Door-Locking Issues
Tesla has announced a massive recall of roughly 1.6 million vehicles in China. This move stems from concerns over two separate issues: potential misuse of the Autopilot driver-assistance system and malfunctioning door-locking mechanisms.
The Chinese regulator, the State Administration for Market Regulation (SAMR), raised the red flag on Autopilot, highlighting the possibility of misuse that could increase the risk of collisions. This closely mirrors the concerns that prompted a recent recall of nearly 2 million Tesla vehicles in the United States. To address this issue, Tesla will employ over-the-air software updates to modify Autopilot's functionality, eliminating the need for time-consuming visits to dealerships or garages. Read More ยปยปยป
The Sunday Brew by The Percolator brings to you curated news on tech, business & entrepreneurship, from across the internet to give your week a perfect start.
Share your thoughts and opinions on the topics covered in this newsletter by leaving a comment and joining the conversation.