Technology

This startup thinks slime mold can help us design better cities

Source

MIT Technology Review

Published

Oct 17, 2025

TL;DR

AI Generated

A startup called Mireta aims to use slime mold's biological abilities to create algorithms that could enhance city design, transit efficiency, and infrastructure resilience. By mimicking how slime mold efficiently distributes resources through networks, Mireta believes its software can optimize various urban systems. Despite skepticism from some experts about the practicality of these algorithms in real-world urban planning, Mireta sees potential in nature-inspired solutions for modern urban challenges. The startup, based in Cambridge, Massachusetts, is also exploring algorithms inspired by ants to further enhance network optimization.

Read Full Article

The robots who predict the future

Three books explore the impact of predictive algorithms on our lives and society. "The Means of Prediction" by Maximilian Kasy delves into how AI predictions shape our decisions and lives, often driven by profit motives. Benjamin Recht's "The Irrational Decision" critiques the reliance on mathematical rationality in computer decision-making, advocating for the value of human intuition. Carissa Véliz's "Prophecy" discusses the power dynamics behind predictions and how they can shape reality. These books collectively highlight the need for greater control over predictive technologies and a reevaluation of their societal implications.

MIT Technology Review•

3 months ago

Hacker News

Algorithms for Optimization [pdf]

The article appears to be a PDF file containing raw data rather than a text-based article. It includes various objects and streams that seem to represent algorithms for optimization. The content is not readable in its current form and requires decoding or conversion to a more accessible format for understanding.

Hacker News•

6 months ago

A Decade Of Architectural RowHammer Defense Solutions (Meta, SNU, UIUC)

Researchers from Meta, Seoul National University, and University of Illinois at Urbana-Champaign published a technical paper titled “SoK: Systematizing a Decade of Architectural RowHammer Defenses Through the Lens of Streaming Algorithms.” The paper addresses the ongoing challenges posed by RowHammer attacks and the need for effective defense solutions. It provides a taxonomy encompassing 48 different works on architectural RowHammer defenses and maps them to streaming algorithms. The researchers offer two practitioner guides to aid in selecting the most suitable defense algorithms based on various factors. The paper also introduces new concepts like Reservoir-Sampling and StickySampling to enhance RowHammer defenses.

SemiEngineering•

6 months ago

Hacker News

Modern Perfect Hashing

The article discusses modern perfect hashing for strings, aiming to map a set of known strings to predefined integers efficiently. The author shares their implementation insights, focusing on optimizing the process for a fixed set of strings. They explore techniques like PEXT and magic values to achieve collision-free mapping. The article delves into code examples for different string lengths and explains the process of finding optimal magic values. The author highlights the importance of efficient splits and the trade-offs between speed and code size. Ultimately, the author achieved a runtime twice as fast as gperf with half the code size, emphasizing the need for a more modern approach in perfect hashing techniques.