Toggle light / dark theme

Artificial general intelligence, or AGI, has become a much-abused buzzword in the AI industry. Now, Google DeepMind wants to put the idea on a firmer footing.

The concept at the heart of the term AGI is that a hallmark of human intelligence is its generality. While specialist computer programs might easily outperform us at picking stocks or translating French to German, our superpower is the fact we can learn to do both.

Recreating this kind of flexibility in machines is the holy grail for many AI researchers, and is often speculated to be the first step towards artificial superintelligence. But what exactly people mean by AGI is rarely specified, and the idea is frequently described in binary terms, where AGI represents a piece of software that has crossed some mythical boundary, and once on the other side, it’s on par with humans.

Last year Swiss Re and Waymo launched a research partnership to define a standard for assessing the risk of autonomous vehicles. One year after that announcement, they are publishing a study that uses real-world data to compare the safety performance of autonomous vs human-driven vehicles. Notably, this is the first time that a robust and significant liability claims dataset is being used to compare the safety performance of autonomous and human drivers.

In fact, Swiss Re was able to produce mileage-and zip-code-calibrated (human driver) private passenger vehicle baselines, against which Waymo’s third party liability insurance claims data were compared. Swiss Re’s baselines, for the specific areas considered, are extremely significant, as they come from over 600,000 claims and over 125 billion miles of exposure.

The results of the research are exciting both for the insurance industry and the safety community alike: in over 3.8 million miles driven without a human being behind the steering wheel in rider-only mode, the Waymo Driver (Waymo’s fully autonomous driving technology) incurred zero bodily injury claims in comparison with the human driver baseline of 1.11 claims per million miles. The Waymo Driver also significantly reduced property damage claims to 0.78 claims per million miles in comparison with the human driver baseline of 3.26 claims per million miles.

Backlash on using AI to redub video game lines.


Naruto’s latest fighting game faces criticism for some questionable voiceover lines, leading to accusations of AI manipulation.

Naruto X Boruto: Ultimate Ninja Storm Connections is the newest brawler based on the hugely popular anime series (which itself adapts the hugely popular manga). Rather than purely adapting one part of the anime, however, Naruto X Boruto skims through almost the entire saga and adapts events already covered in previous games. But when comparing the new game’s dub to those previous entries, fans have been left confused.

A research team led by Prof. Chen Wei from the University of Science and Technology of China (USTC) of the Chinese Academy of Sciences (CAS) designed a rechargeable hydrogen-chlorine (H2-Cl2) battery that can operate in temperatures ranging from −70°C to 40°C. The study was published in Journal of the American Chemical Society as the cover article.

Hydrogen fuel cells are a promising technology valuable for their sustainability and the abundance of hydrogen, among which H2-Cl2 fuel cells stand out due to fast electrochemical kinetics, high redox potential and high specific capacity of Cl2/Cl- redox couple. However, the volatile chlorine gas cannot be retained during the charging process, resulting in poor Coulombic efficiency (CE) and reversibility. There is an urgent need to develop aqueous chlorine batteries with and applicability at different temperatures.

The research team first discovered that due to the lack of binding sites with strong affinity to Cl2, traditional adsorptive cathodes have difficulty immobilizing Cl2, causing low reversibility. To tackle this problem, the team designed a hierarchically porous carbon composed of highly micro-/mesoporous carbon (HPC) and macroporous carbon felt (CF), effectively confining the Cl2 on the cathode and improving the reversibility.

Hilary Putnam has argued that computational functionalism cannot serve as a foundation for the study of the mind, as every ordinary open physical system implements every finite-state automaton. I argue that Putnam’s argument fails, but that it points out the need for a better understanding of the bridge between the theory of computation and the theory of physical systems: the relation of implementation. It also raises questions about the class of automata that can serve as a basis for understanding the mind. I develop an account of implementation, linked to an appropriate class of automata, such that the requirement that a system implement a given automaton places a very strong constraint on the system. This clears the way for computation to play a central role in the analysis of mind.

https://aperture.gg/information.
Visit https://www.acorns.com/Aperture to get a $10 bonus when you sign up to grow your oak!
T&C’s apply. Merch: https://aperture.gg/merch.

The modern age of information is possible thanks to the work of a single person, one who changed the way we viewed the world; most people have no idea.

To Claude Shannon, thank you.

Stay connected with Aperture: