Skip to main content

Climate Q&A by Ekimetrics, AI with a purpose

 


Hugging Face, Ekimetrics  and Gradio an example of connected narrow AI at it's most useful

Ekimetrics describes itself as "a pioneering leader in data science and AI-powered solutions for sustainable business performance." I've just discovered their Climate Q&A on Hugging Face. Put simply it's the best tool I've come across, by far, for interrogating the IPCC reports. 

Hugging Face

In case you haven't come across Hugging Face it's a company that develops tools to help people build machine learning applications. Its most notable product is the Transformers library, which is a collection of pre-trained machine learning models for natural language processing tasks. Hugging Face also operates a platform that allows users to share machine learning models and datasets.

Hugging Face's products and services:

  • Transformers library: The Transformers library is a collection of pre-trained machine learning models for natural language processing tasks. These models can be used to perform a variety of tasks, such as text classification, text generation, and question answering.
  • Hugging Face Hub: The Hugging Face Hub is a platform that allows users to share machine learning models and datasets. The Hub is a valuable resource for researchers and developers who are working on machine learning projects.
  • Hugging Face is a leading provider of tools for building machine learning applications. Its products and services are used by researchers, developers, and businesses around the world.

Climate Q&A 🌍

Ask climate-related questions to the IPCC reports at https://huggingface.co/spaces/Ekimetrics/climate-question-answering

Ekimetrics 

'We help companies get more from their data and implement pre-packaged AI solutions, so they can combine high impact with long-term business purpose.

Information regarding climate change can be complex to interpret. In order to bridge the gap to make science more accessible, Climate Q&A allows you to ask questions about the @IPCC_CH  reports and easily understand the answers.' 

The Query

You can see the query I tested Climate Q&A with in the screenshot below. What was impressive was the detailed referencing provided, as can be seen



The answer given by Climate Q&A was: 

Based on the available documents, there is no specific information on how the 2023 El Niño is expected to affect European temperatures. However, the following information can be inferred:

- Extreme El Niño events are projected to occur about twice as often in the 21st century compared to the 20th century, and they are likely to intensify existing hazards, with drier or wetter responses in several regions across the globe [docs 2, 3].

- With increasing warming, the frequency, intensity, and duration of heat-related events, including heatwaves, are projected to continue to increase through the 21st century, particularly in the Mediterranean region and southern Africa [doc 1].

- Mean and maximum temperatures, frequencies of warm days and nights, and heatwaves have increased since 1950, while the corresponding cold indices have decreased with average warming projected to be larger than the global mean in all of Europe, with the largest winter warming in NEU and EEU and the largest summer warming in MED [doc 4].

- Extreme precipitation is projected to increase in both boreal winter and summer over Europe, with regional differences, and the most intense precipitation events observed today in Europe are projected to almost double in occurrence for each 1degC of further global warming [docs 5, 9].

- Warming in Europe will continue to rise faster than the global mean, widening risk disparities across Europe in the 21st century, with largely negative impacts projected for southern regions and some short-term benefits anticipated in the north [docs 7].

- There is high confidence that most coastal/ocean-related climatic impact-drivers in Europe will increase over the 21st century for all scenarios and time horizons, with relative sea level rise extremely likely to continue around Europe, contributing to increased coastal flooding in low-lying areas and shoreline retreat along most sandy coasts [doc 8].

- Projections for the 21st century show increases in storms across all of Europe for >2degC GWL with a decrease in their frequency in the MED [doc 10].

In summary, while there is no specific information on the impact of the 2023 El Niño on European temperatures, it is likely that Europe will continue to experience warming temperatures, with an increase in extreme weather events such as heatwaves and extreme precipitation. The Mediterranean region is expected to be particularly affected, with negative impacts projected for southern regions. 

This isn't all, plenty of further, useful in formation is available, including a full list of references.


Climate Q&A was built by built using the Gradio app 'the fastest way to demo your machine learning model with a friendly web interface so that anyone can use it, anywhere!'

All round it's very impressive. To add the cherry on top of the cake, Climate Q&A reported that: "one question asked to ClimateQ&A is around 0.482gCO2e - equivalent to 2.2m by car (https://datagir.ademe.fr/apps/impact-co2/) Or around 2 to 4 times more than a typical Google search."

Which begs a question, a big one: how large will the CO2e increase with due to the popularity of AI tools, in the next year? 



Labels:

Comments

Post a Comment

Popular posts from this blog

The AI Dilemma and "Gollem-Class" AIs

From the Center for Humane Technology Tristan Harris and Aza Raskin discuss how existing A.I. capabilities already pose catastrophic risks to a functional society, how A.I. companies are caught in a race to deploy as quickly as possible without adequate safety measures, and what it would mean to upgrade our institutions to a post-A.I. world. This presentation is from a private gathering in San Francisco on March 9th with leading technologists and decision-makers with the ability to influence the future of large-language model A.I.s. This presentation was given before the launch of GPT-4. One of the more astute critics of the tech industry, Tristan Harris, who has recently given stark evidence to Congress. It is worth watching both of these videos, as the Congress address gives a context of PR industry and it's regular abuses. "If we understand the mechanisms and motives of the group mind, it is now possible to control and regiment the masses according to our will without their
Post a Comment
Read more

Beware the Orca, the challenge to ChatGPT and Palm2 is here

  So Google's 'we have no moat' paper was correct. If you train an LLM wisely then it's cost effective and cheap to produce a small LLM that is able to compete or even beat established, costly LLMs, as Microsoft has just found. It's another excellent video from AI Explained, who goes through some of the training procedures, which I won't get into here. Orca, is a model that learns from large foundation models (LFMs) like GPT-4 and ChatGPT by imitating their reasoning process. Orca uses rich signals such as explanations and complex instructions to improve its performance on various tasks. Orca outperforms other instruction-tuned models and achieves similar results to ChatGPT on zero-shot reasoning benchmarks and professional and academic exams. The paper suggests that learning from explanations is a promising way to enhance model skills. Smaller models are often overestimated in their abilities compared to LFMs, and need more rigorous evaluation methods. Explana
Post a Comment
Read more

What is happening inside of the black box?

  Neel Nanda is involved in Mechanistic Interpretability research at DeepMind, formerly of AnthropicAI, what's fascinating about the research conducted by Nanda is he gets to peer into the Black Box to figure out how different types of AI models work. Anyone concerned with AI should understand how important this is. In this video Nanda discusses some of his findings, including 'induction heads', which turn out to have some vital properties.  Induction heads are a type of attention head that allows a language model to learn long-range dependencies in text. They do this by using a simple algorithm to complete token sequences like [A][B] ... [A] -> [B]. For example, if a model is given the sequence "The cat sat on the mat," it can use induction heads to predict that the word "mat" will be followed by the word "the". Induction heads were first discovered in 2022 by a team of researchers at OpenAI. They found that induction heads were present in
Post a Comment
Read more