When universities speak of “campus growth,” what kind of growth do they mean? More students? (At the moment, international students comprise about 5 percent of total enrollment) More research? (61 percent of graduate students were international) More healthcare? (Healthcare enterprises are moving off campus to where the patients are). The traditional campus concentrated infrastructure — the modern academic health system distributes it. (At what point does a university cease to be a campus and become a regional enterprise?) More computing? (Electric or absorption chillers). Or are they simply replacing infrastructure that has reached the end of its useful life?
Data centers in colleges and universities are crucial for supporting the extensive technological infrastructure required for modern education and research. These centers house critical servers and storage systems that manage vast amounts of data, ensuring reliable access to academic resources, administrative applications, and communication networks. They enable the secure storage and processing of sensitive information, including student records, faculty research, and institutional data.
Moreover, data centers facilitate advanced research by providing the computational power needed for data-intensive studies in fields like bioinformatics, climate science, and artificial intelligence. They support virtual learning environments and online course management systems, essential for the increasingly prevalent hybrid and online education models. Efficient data centers also contribute to campus sustainability goals by optimizing energy use through modern, eco-friendly technologies.
Additionally, robust data center infrastructure enhances the university’s ability to attract top-tier faculty and students by demonstrating a commitment to cutting-edge technology and resources. They also play a vital role in disaster recovery and business continuity, ensuring that educational and administrative functions can resume quickly after disruptions. Overall, data centers are integral to the academic mission, operational efficiency, and strategic growth of colleges and universities.
Standards distinguish meaningful patterns from imagined ones.
We have followed development of the technical standards that govern the success of these “installations” since 1993; sometimes nudging technical committees — NFPA, IEEE, ASHRAE, BICSI and UL. The topic is vast and runs fast so today we will review, and perhaps respond to, the public consultations that are posted on a near-daily basis. Use the login credentials at the upper right of our home page.
Abstract: Power grid operation and maintenance decision-making reform is an important part of power system reform. With the construction of massive historical quasi real-time data management platform, the reform of power system is also advancing. However, in the face of massive data explosion, the business level and business logic become disorganized and redundant. Based on the actual situation of Shenzhen Power Supply Bureau, the sg-erp data center is composed of structured data center, massive data center, unstructured data center and power grid GIS data center. With the unprecedented growth of business application data, the data center can improve business logic and promote power system reform. The experimental results show that big data technology has a broad application prospect in the reform of power industry.
American English is effectively the de facto reference language for most modern LLM tokenization. During today’s session we explore the at-present advantage Americans have in the development of artificial applications — whether it should always be that way or not. Tokenization isn’t language-neutral — it’s heavily skewed toward English due to data realities. This is one of the core reasons why “English-first” prompting often works best in today’s LLMs.
We will use the document linked below to begin the exploration:
Final Input to the AI Model:
[464, 2068, 7583, 1776, 18045, 625, 262, 16925, 3290, 13]
Background:
The model only sees this list of numbers. It has no direct understanding of English words anymore — it learned patterns from billions of examples during training using these number sequences.This numerical input then goes through embeddings (turning numbers into vectors), attention layers, etc., to generate a response. Most widely used tokenizers (e.g., OpenAI’s tiktoken, Llama’s, etc.) are trained primarily on English-heavy datasets (often 60–90%+ English in pre-training corpora).
Outcome:
Better compression for English — Common English words and patterns become single tokens or short subwords.
Worse efficiency for other languages — Non-English text often gets fragmented into more tokens (sometimes 2–5× more for the same semantic content).
Impact:
Higher token counts = higher API costs and shorter effective context windows for non-English users.
Poorer downstream performance on non-English tasks.
English becomes the “cheapest” and often “best-performing” language for prompting and reasoning.
Studies consistently show this “tokenization tax” or “language premium”: English typically has the lowest token-per-character or token-per-meaning ratio in major models.
Bias:
Multilingual models still underperform on low-resource languages.
It reinforces English as the default language for AI development.
It affects fairness, accessibility, and global adoption.
Efforts to fix this include dedicated multilingual tokenizers, language-specific fine-tuning, and more balanced approaches. However, because English dominates training data and benchmarks, it remains the practical standard that everything else is measured against.
Tokenization isn’t language-neutral — it’s heavily skewed toward English due to data realities. This is one of the core reasons why “English-first” prompting often works best in today’s LLMs.
* StandardsMichigan.COM normally deals with Language issues every Monday at least once per month.
The International Trade Administration (ITA) of the U.S. Department of Commerce (DOC) is requesting public comments to gain insights on the current global artificial intelligence (AI) market. Responses will provide clarity about stakeholder concerns regarding international AI policies, regulations, and other measures which may impact U.S. exports of AI technologies. Additionally, the request for information (RFI) includes inquiries related to AI standards development. ANSI encourages relevant stakeholders to respond by ITA’s deadline of October 17, 2022.
They put in new turf and then immediately covered it with thousands of pounds of equipment and rigging for the Morgan Wallen concert? https://t.co/B37WCDpVHx
“Whom the gods love, die young.” Menander (341-290 BCE)
“My mother was a Protestant, of a traditional American, vague kind: she belonged to the church that the nice people in the neighbourhood went to. My wife is a Catholic, the kids are Catholic, so I’m a Catholic fellow-traveller.”
“When buying and selling are controlled by legislation, the first things to be bought and sold are legislators.”
“If government were a product, selling it would be illegal.”
“The Democrats are the party that says government will make you smarter, taller, richer, and remove the crabgrass on your lawn. The Republicans are the party that says government doesn’t work and then they get elected and prove it.”
“Not much was really invented during the Renaissance, if you don’t count modern civilization.”
“No humorist is under any obligation to provide answers and probably if you were to delve into the literary history of humour it’s probably all about not providing answers because the humorist essentially says: this is the way things are.”
“Panoramic View of the Greek Amphitheatre at Syracuse” 18th Century Abraham Louis Rodolphe Ducros
From from time to time — particularly in the months of fairer weather, when many events are hosted outdoors — we break form from the grind of responding to c0nsultations to simply enjoy these spaces See our CALENDAR for our periodic Lively c0lloquia when we drill down into technical specifics.
서울대학교
Outdoor amphitheaters rely upon a broad ecosystem of consensus standards that quietly support every performance. Among them:
NFPA — National Electrical Code, including Articles 518, 520, 525 and related provisions
SMPTE — media production and broadcast systems
Audio Engineering Society (AES) — professional sound systems
Illuminating Engineering Society (IES) — site and performance lighting
ASTM International — seating, concrete, accessibility and materials
ASCE — structural loading for stages, roofs and canopies
“Panoramic View of the Greek Amphitheatre at Syracuse” / Abraham Louis Rodolphe Ducros (18th Century)
Today at 15:00 UTC explore best practice literature for outdoor events in cross-cutting fashion; refreshing our understanding about how we make such events in academic settings safe, sustainable and successful. Use the login credentials at the upper right of our home page.
Hosting open-air celebrations such as graduations and cultural requires sensitivity to audio standards that ensure high-quality sound and audience safety. Today at the usual hour we examine the standards covering sound system design, noise control, loudness levels, equipment specifications, weather and local environmental considerations.
Surprisingly, there are many even without approaching the not insignificant cabling system standards. We pick three representatives titles which cross reference one another:
World Health Organization recommendations for limiting exposure to leisure noise (e.g., concerts) to a level exceeding 80 dBA for 24 hours or 140 dB peak sound pressure to prevent hearing loss. Recommendations for speech heavy events will defer from musical events.
Height and Angle: Line array speakers are often elevated (e.g., on trusses or poles) and angled to cover the audience while reducing sound spill to non-audience areas.
Distance: Main speakers should be placed to avoid excessive sound pressure near the front (e.g., 95–100 dBA max for audience safety) and ensure clarity at the back.
Acoustical Society of America ASA-2010 Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools. While primarily for schools, this standard’s principles can apply to open-air educational or community events, emphasizing background noise control and speech intelligibility. For open-air audiences at events like outdoor lectures or festivals, speaker placement should minimize interference from environmental noise (e.g., traffic, wind). The standard suggests maintaining a signal-to-noise ratio where speech is at least 15 dB above background noise for clarity.
Use the login credentials at the upper right of our home page.
New update alert! The 2022 update to the Trademark Assignment Dataset is now available online. Find 1.29 million trademark assignments, involving 2.28 million unique trademark properties issued by the USPTO between March 1952 and January 2023: https://t.co/njrDAbSpwBpic.twitter.com/GkAXrHoQ9T
