Home

I first became aware of GPT-3 technology in late spring or summer of 2020. San Francisco-based OpenAI engineers and researchers were comparing GPT-3’s capacity to that of its predecessor.(Brown 2020:14) CNBC described the lab as backed by Elon Musk.(Shead 2020)

On 17 January 2023, an 11-year old boy showed me his Q&A in the “playground” version. Within minutes he had the app installed on my PC along with its companion for images.

The first time I used it was to reword a sentence in a document I was citing. I experimented with a number of questions, voted thumbs up or down on answers, provided feedback, reworded questions, and checked search engines for the reliable sources Chat does not provide. It was a bit like that feeling the first time you experiment with HTML or a search engine.

I asked GPT-Chat to reword the phrase, “the sector grew a third faster than the Canadian economy as a whole” which was the exact wording of the original document.. The app suggested, the sector “experienced a growth rate that was one-third higher than the overall economy”. I published the result at 1:44 on 18 January 2023. I later changed it to “The sector experienced a growth rate that was one-third higher than Canada’s overall economy from 2010 to 2017.”

  • Me: All environmental and clean technology activities accounted for what percentage of Canada’s gross domestic product in 2019?
  • GPT-3 Chat Answer: I’m sorry, I don’t have the exact percentage of Canada’s GDP that environmental and clean technology activities accounted for in 2019. However, I can tell you that the clean technology sector has been growing in recent years and is considered a key part of Canada’s economy. According to the Canadian clean technology industry report 2020, the clean technology sector has grown by 4.5% annually over the past five years, and the country’s clean technology exports reached $10.5 billion in 2019. The Canadian government has also been investing in clean technology and renewable energy, and has set targets to reduce greenhouse gas emissions and transition to a low-carbon economy.

I entered “Canadian clean technology industry report 2020 $10.5 billion in 2019” in BRAVE’s search engine which resulted in a link to a 29 January 2020 Business Insider article entitle, “Private investors poured $10.5 billion into clean energy in 2019. But this is not specific to Canada.

Key words “Canada’s clean technology exports reached $10.5 billion in 2019” search results led to Statistics Canada’s, The Daily Annual Survey of Environmental Goods and Services, 2019 which said that, “Canadian businesses sold $18.9 billion worth of environmental and clean technology goods in 2019, while sales of related services amounted to $10.2 billion. If waste management services are included, the total revenue from sales of environmental and clean technology services rises to $18.5 billion.”

When asked to reword a sentence, GPT-3 Chat generated “renewable energy refers to energy sources that are replenished naturally and at a rate that is equal to or faster than the rate at which they are utilized.” I later found this sentence in a NRCAN document, “Renewable energy is derived from natural processes that are replenished at a rate that is equal or faster than the rate at which they are consumed.”

I noticed that GPT-3 adds more details in its suggested rewording of a phrase and includes things that were not in the original text, making it inaccurate and unreliable. But if you search those suggestions, one by one, to find reliable sources, it can be helpful in developing a topic or providing answers to questions. Is it faster or better than a search engine?

It is particularly easy to find reliable sources when the app provides numerical data with dates, etc.

Some interesting outcomes in my first 24 hours with the technology, is the way in which timelines generated by GPT-3 technology includes dates of projected goals: “2050: The government announces that Canada has achieved net-zero emissions.”

It is a bit like the bots operated by humans that perform a multitude of mundane editing corrections. I thank them every time, sometimes just because it is so helpful that others are improving something I am working on without even being asked.

I feel like I want to thank the OpenAI GPT-3 Chat every time its answers are really useful, even if they are not ready for publication. I want to give it a name, like you would a pet dog or cat.

Even that makes me pause.

Some of the concerns raised and criticism related to the technology are valid. The implications are impossible to imagine at this point.

But is isn’t going to go away.

The free playgrounds may become fenced in. With a paywall surrounding it, many of us may be on the outside peeking in. Some of us will feel a profound sense of loss if and when that happens.

So for now, I want to take full advantage of this free technology. I will have to be extra attentive to the seductive ways in which I can be lulled into accepting text that is not quite accurate as a short cut for producing content. The price would be that my future self on reading the content, if left unchallenged and unchanged, would also be deceived by what my present self allowed to become published and public.

Wikipedia in itself is not a reliable source. When inline citations to reliable sources are provided, the encyclopedia can help answer some questions and help the reader generate more questions. Wikipedia and OpenAI GPT-3 Chat can be useful tools in the lifelong learning process, as long as the human mind never ceases to engage in critical thinking, to ask the questions, Who wrote that and when? What sources informed their knowledge claim? How do we know they are reliable and up-to-date? In the case of GPT-3 technologies, the question changes to, “What are the reliable sources in the genealogies of knowledge claims that GPT-3 generated?” It is a bit like reverse engineering.

The 1999 Bahá’í International Community (BIC) document on the state of human society, Who Is Writing the Future?, said that access to education to the masses in countries had accelerated in the twentieth century. In the international community, the “World Bank, government agencies, major foundations and several branches of the United Nations system” contributed to this acceleration. The authors said that the second significant factor to advancement in learning was the explosion in “information technology”. These new technologies have “made all of the earth’s inhabitants potential beneficiaries of the whole of the race’s learning.”

References


  • Under construction

Memory work is a “process of engaging with the past which has both an ethical and historical dimension”.[1]

The 1619 Project was a “major initiative” of the The New York Times that was inaugurated on 18 August 2019 with the special issue of The New York Times Magazine.[4:4] It was published 400 years after the arrival of the first enslaved people from Ghana, Africa to Point Comfort, Virginia, then a British Colony.[4:4 The Project sought to reframe the history of the United States by centering the “consequences of slavery and the contributions of Black Americans” in the US “national narrative.”[3]

The first essay in the series is an introduction by Times staff writer, Nikole Hannah-Jones, who was the inspiration behind the project.

‘The following is a list of contributions to the 1619 Project that were published in the Times Magazine along with links to related Wikipedia articles.

  • Introduction and intellectual framework by Nikole Hannah-Jones Pages 14 – 22
  • 400 Years: A Literary Timeline
    • “August 1619” a poem about the Middle Passage by Clint Smith. Page 28
      • Smith’s poem is the first in a series of 16 stories and poems by writers who described 400 years of “consequential moments in African-American history” In “August 1619”which was published by the Times Magazine on 14 August 2019Smith describes how he is “chasing a history that swallowed me” as he traces lines on a globe from the coast of Angola to Brazil, from Ghana to Jamaica, and from Senegal to South Carolinawhere 36,000 slave ships crossed the Atlantic Ocean over 350 years.[3][See Notes] See the related Middle Passage and the archived copy of the 100-page 1619 Project PDF here.[4]
  • “African & Natick blood-born…” a poem about Crispus Attucks by Yusef Komunyakaa p.29
    • Attucks, who died on March 5, 1770, fighting against the British in the “Boston Massacre” during the American Revolution, was the first American to die in the Revolution. He was a dock worker of  Wampanoag and African descent, who had escaped from slavery.
  • 1619 Project:42 Eve L. Ewing on Phillis Wheatley Wikipedia article
  • 1619 Project:43 Reginald Dwayne Betts on the Fugitive Slave Act of 1793 Wikipedia article
  • 1619 Project:46 Barry Jenkins on Gabriel’s Rebellion Wikipedia article
  • 1619 Project:47 Jesmyn Ward on the Act Prohibiting Importation of Slaves Wikipedia article
  • 1619 Project:58 Tyehimba Jess on Black Seminoles Wikipedia article
  • 1619 Project:59 Darryl Pinckney on the Emancipation Proclamation of 1863 Wikipedia article
  • 1619 Project:59 ZZ Packer on the New Orleans massacre of 1866 Wikipedia article
  • 1619 Project:68 Yaa Gyasi on the Tuskegee Syphilis Study see Wikipedia article
  • 1619 Project:69 Jacqueline Woodson on Sgt. Isaac Woodard Wikipedia article
  • 1619 Project:78 Rita Dove and Camille T. Dungy on the 16th Street Baptist Church bombing Wikipedia article
  • 1619 Project:79 Joshua Bennett on the Black Panther Party Wikipedia article
  • 1619 Project:84 Lynn Nottage on the birth of hip-hop Wikipedia article
  • 1619 Project:84 Kiese Laymon on the Rev. Jesse Jackson’s “rainbow coalition” speech Wikipedia article Wikiquote 1984 Rainbow speech no specific article yet
  • 1619 Project:85 Clint Smith on the Superdome after Hurricane Katrina Wikipedia article
  • Origins of American capitalism. Essay by Mehrsa Baradaran. Pages 31-40.

Notes

  • Eltis, David. 31 March 2020. “Digital Resources: The Trans-Atlantic Slave Trade Database.” Oxford Research Encyclopedias. Oxford University. https://doi.org/10.1093/acrefore/9780199366439.013.906 From “a database of 36,000 slave-trading voyages between Africa and the New World, and second, a database of 11,400 voyages from one port in the Americas to another—a traffic known as the intra-American slave trade. The time span covered is from the 16th to the late 19th century. The site also offers personal information on 92,000 Africans found on board some of those voyages, which is stored in a separate database, as well as an interface that permits users to explore our estimates of the overall size and direction of the transatlantic slave trade broken down by each of the 340 years of its existence.” This database is available through SlaveVoyages.org.

References

  1. Memory Work“. Wikipedia. accessed 14 January 2023.
  2. Nikole Hannah-Jones. Wikipedia. accessed 14 January 2023.
  3. Bessner, Daniel. 14 January 2023. “Americans Are Fighting Over History While Historians Disappear.” The New York Times.
  4. Smith, Clint. 14 Aug 2019. “August 1619”. A New Literary Timeline of African-American History. Archived.
  5. Hannah-Jones, Nikole. 18 August 2019. The 1619 Project. The Pulitzer Centre. 100 pp. Archived PDF url

Wikipedia articles cited

Webcams

July 13, 2020


Broadcast is powered by Baltic Live Cam

 


Why I chose Roam Research

I would sometimes see visualizations relating aspects of Wikipedia articles and imagine what my own research would resemble as an image. I knew it would not be tree-like with a trunk and branches but like grass with rhizome roots and sister connections. The first time I saw the “graph” of my first 3 days on Roam Research, I came close to getting teared up.

Working as Wikipedia editor – a veteran volunteer – has changed my mind. As more tools became available to me, such as interconnectivity with Zotero, I seek them whenever I read. Roam has the same feature. Ctrl+Shift+C creates a template that can be embedded in both saving tons of work in the long run. The template generates any number of reference styles.

Rather than trying to figure out which folder/subfolder to place a page/note/topic, the square brackets generate the link to that item. In Wikipedia, I can choose from available categories and/or portals or create new ones, in Roam I am sure there is a similar option.

Roam autocomplete automatically suggests notes I have already written which also saves time.

Roam has a to do list.

Other options

  • Microsoft OneNote – $0
  • Google Keep – Google users
  • Joplin “replacement for Evernote”; “open source note taking and to-do application” with synchronisation” capabilities
  • Notion: $0 then $4/m+ – Teamwork
  • Org-mode Wikipedia article
  • Turtl
  • Simplenote distraction-free
  • Standard Notes secure, encrypted note taking
  • Evernote Basic (free), Plus ($34.99 per year), and Premium ($69.99 per year).
  • Roam Research
  • Federated Wiki
  • Boostnote Developers
  • Supernotes
  • Zettlr
  • Draft
  • TiddlyWiki
  • Workflowy
  • Dynalist
  • Zim
  • Quip $12/m
  • Trello
  • Asano

MSRP Manufacturer’s Suggested Retail Price or retail price


I began to appreciate these technologies while updating some Wikipedia articles of the wildfires in Alberta and then in the Amazon. The excerpts below are based on an March 19 ESA article “COVID-19: nitrogen dioxide over China.”

Some of the content is from the Wikipedia article on the Sentinel-5 Precursor.

The European Space Agency (ESA)’s Earth observation satellite, Sentinel-5 Precursor, which is part of the Copernicus Programme is monitors and reports on air pollution. The Tropomi instrument on the 5P satellite, showed a “dramatic reduction in nitrogen dioxide concentrations” in “all major Chinese cities between late-January and February,” 2020. China had enacted strict measures in late December 2019, and by late January, daily activities in China had ceased following orders by Chinese authorities to close and to clear streets. According to the European Space Agency (ESA) Nitrogen dioxide is released by “power plants, industrial facilities and vehicles.”

Studies, based on Copernicus Atmosphere Monitoring Service’s (CAMS) combination of “satellite observations with detailed computer models of the atmosphere, said that, in February 2002, compared to the three previous years, there was a decrease of about 20 to 30 percent one of the “most important air pollutants”—fine suspended particulate matter (SPM) over large parts of China.”

The March 19 article also said that recent data showed a “decline of air pollution over northern Italy coinciding with its nationwide lockdown to prevent the spread of the coronavirus (COVID-19).”

The European_Space_Agency (ESA)’s Earth observation satellite, Sentinel-5 Precursor, which is part of the Copernicus Programme is monitoring and reporting on air pollution.

“The Tropomi instrument on the 5P satellite, showed a “dramatic reduction in nitrogen dioxide concentrations” in “all major Chinese cities between late-January and February,” 2020. China had enacted strict measures in late December 2019, and by late January, daily activities in China had ceased following orders by Chinese authorities to close and to clear streets. According to the European Space Agency (ESA) Nitrogen dioxide is released by “power plants, industrial facilities and vehicles.”

“Studies, based on Copernicus Atmosphere Monitoring Service’s (CAMS) combination of “satellite observations with detailed computer models of the atmosphere, said that, in February 2002, compared to the three previous years, there was a decrease of about 20 to 30 percent one of the “most important air pollutants”—fine suspended particulate matter (SPM) over large parts of China.”

Notes

The “Tropomi (TROPOspheric Monitoring Instrument) is a spectrometer sensing ultraviolet (UV), visible (VIS), near (NIR) and short-wavelength infrared (SWIR) to monitor ozone, methane, formaldehyde, aerosol, carbon monoxide, NO2 and SO2 in the atmosphere.”