Archives: Bookmarks
The Use of Automatic AI-based Notes and Transcription Services in Qualitative Research: Ethical and Methodological Concerns | Proceedings of the ALISE Annual Conference
Ethical and methodological concerns surrounding AI-based meeting assistants’ use were outlined, including data ownership and intellectual property management, data repository and flow strategies, real-time nature of interactions, implicit speech bias, and decision-making accountability. The study contributes to broadening the understanding of the ethical and methodological concerns for qualitative researchers when using AI meeting assistants and further calls for specific LIS guidelines in online research using AI technology.
DeepSeek – Into the Unknown
DeepSeek | DeepSeek-V2 | Kaggle
“DeepSeek Prover is an improved open-source language model
for theorem proving in Lean 4. It offers optimized training and
inference processes with state-of-the-art results across multiple
benchmarks.”
[2404.12534] Towards Large Language Models as Copilots for Theorem Proving in Lean
Theorem proving is an important challenge for large language models (LLMs), as formal proofs can be checked rigorously by proof assistants such as Lean, leaving no room for hallucination. Existing LLM-based provers try to prove theorems in a fully autonomous mode without human intervention. In this mode, they struggle with novel and challenging theorems, for which human insights may be critical. In this paper, we explore LLMs as copilots that assist humans in proving theorems. We introduce Lean Copilot, a framework for running LLM inference in Lean. It enables programmers to build various LLM-based proof automation tools that integrate seamlessly into the workflow of Lean users. Using Lean Copilot, we build tools for suggesting proof steps (tactic suggestion), completing intermediate proof goals (proof search), and selecting relevant premises (premise selection) using LLMs. Users can use our pretrained models or bring their own ones that run either locally (with or without GPUs) or on the cloud. Experimental results demonstrate the effectiveness of our method in assisting humans and automating theorem proving process compared to existing rule-based proof automation in Lean. We open source all codes under a permissive MIT license to facilitate further research.
Mathematicians’ Newest Assistants Are Artificially Intelligent | Scientific American
If you are writing a proof with the help of the Caltech co-pilot, you can click on a button to request new lines of Lean’s programming language to represent the mathematics you are working with. Several options, which Anandkumar calls “tactic suggestions,” will appear on the right-hand side of the screen; you then simply choose whichever option looks most appropriate. If your proof is headed in a direction that has obvious or well-known intermediate conclusions, the co-pilot can also suggest how to complete that trajectory.
“There’s no trust issue” with Lean because the software checks the work, says Martin Hairer, a professor of pure mathematics at the Swiss Federal Institute of Technology in Lausanne and at Imperial College London. Still, many academics haven’t adopted it yet. “It’s hard to use because you have to enter all the mathematics as code,” Hairer says. Coding in Lean requires entering details that would be omitted when writing up a paper, he notes, so it might take multiple pages of code to show what’s self-evidently true or obvious.
llama-models/models/llama3_1/MODEL_CARD.md at main · meta-llama/llama-models
“Training Greenhouse Gas Emissions Estimated total location-based greenhouse gas emissions were 11,390 tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy, therefore the total market-based greenhouse gas emissions for training were 0 tons CO2eq.”
As Use of A.I. Soars, So Does the Energy and Water It Requires – Yale E360
Right now, it’s not possible to tell how your A.I. request for homework help or a picture of an astronaut riding a horse will affect carbon emissions or freshwater stocks. This is why 2024’s crop of “sustainable A.I.” proposals describe ways to get more information about A.I. impacts.
In the absence of standards and regulations, tech companies have been reporting whatever they choose, however they choose, about their A.I. impact, says Shaolei Ren, an associate professor of electrical and computer engineering at UC Riverside, who has been studying the water costs of computation for the past decade. Working from calculations of annual use of water for cooling systems by Microsoft, Ren estimates that a person who engages in a session of questions and answers with GPT-3 (roughly 10 t0 50 responses) drives the consumption of a half-liter of fresh water. “It will vary by region, and with a bigger A.I., it could be more.” But a great deal remains unrevealed about the millions of gallons of water used to cool computers running A.I., he says.
The same is true of carbon.
“Data scientists today do not have easy or reliable access to measurements of [greenhouse gas impacts from A.I.], which precludes development of actionable tactics,” a group of 10 prominent researchers on A.I. impacts wrote in a 2022 conference paper. Since they presented their article, A.I. applications and users have proliferated, but the public is still in the dark about those data, says Jesse Dodge, a research scientist at the Allen Institute for Artificial Intelligence in Seattle, who was one of the paper’s coauthors.
Projecting the Electricity Demand Growth of Generative AI Large Language Models in the US – Center on Global Energy Policy at Columbia University SIPA | CGEP %
Based on this approach, and in light of data from the EIA Annual Energy Outlook 2023,[13] this analysis suggests that by 2027 GPUs will constitute about 1.7 percent of the total electric capacity or 4 percent of the total projected electricity sales in the United States. While this might seem minimal, it constitutes a considerable growth rate over the next six years and a significant amount of energy that will need to be supplied to data centers.
Why AI Consumes So Much Energy and What Might Be Done About It – Kleinman Center for Energy Policy
So when you think about the figures that are out there, most people describe the data center as consuming anywhere from 1 to 2% of electricity that’s consumed globally. And of that, when you think about AI, they say roughly about 12% of that data center footprint today is occupied by AI-based workloads.
So in this context, today the current footprint of AI is roughly about .1 to .2%, depending on the numbers you’re looking at. But I think the thoughts or interest is understanding how we continue to keep up with the demand for what’s happening in a lot of these AI development spaces — not just training, but also for inference — and how we make sure that we can not have the growth outpace our ability to offset the overall impact on the environment and the grid.