DeepMind: [A Generalist Agent](https://www.deepmind.com/publications/a-generalist-agent)
> Abstract
> Inspired by progress in large-scale language modelling, we apply a similar approach towards building a single generalist agent beyond the realm of text outputs. The agent, which we refer to as Gato, works as a multi-modal, multi-task, multi-embodiment generalist policy. The same network with the same weights can play Atari, caption images, chat, stack blocks with a real robot arm and much more, deciding based on its context whether to output text, join...
**[PaLM-E: An Embodied Multimodal Language Model](https://palm-e.github.io/)** by Google
> Large language models have been demonstrated to perform complex tasks. However, enabling general inference in the real world, e.g. for robotics problems, raises the challenge of grounding. We propose embodied language models to directly incorporate real-world continuous sensor modalities into language models and thereby establish the link between words and percepts. Input to our embodied language model are multi-modal sentences that interleave visual, continuous s...
WSJ: [Google DeepMind CEO Says Some Form of AGI Possible in a Few Years](https://www.wsj.com/articles/google-deepmind-ceo-says-some-form-of-agi-possible-in-a-few-years-2705f452)
> Artificial general intelligence, a system in which computers have human-level cognitive abilities, could be achievable within a few years, said Google DeepMind Chief Executive Officer Demis Hassabis.
> Mr. Hassabis, speaking Tuesday at The Wall Street Journal’s Future of Everything Festival, said the speed of AI research could accelerate from its already-rapid pace.
> “The p...
The Hill: [Biden has told Obama he’s running again](https://thehill.com/news/administration/3272281-biden-has-told-obama-hes-running-again/)
> President Biden has told former President Obama that he is planning to run for reelection in 2024, two sources tell The Hill.
> The admission to Obama is the latest indication that Biden is likely to run for a second term, something the president has spoken about publicly.
> During a press conference in Brussels last month, he told reporters he’d be “very fortunate” to run against his rival in the 2020 electio...
WSJ: [Google DeepMind CEO Says Some Form of AGI Possible in a Few Years](https://www.wsj.com/articles/google-deepmind-ceo-says-some-form-of-agi-possible-in-a-few-years-2705f452)
> Artificial general intelligence, a system in which computers have human-level cognitive abilities, could be achievable within a few years, said Google DeepMind Chief Executive Officer Demis Hassabis.
> Mr. Hassabis, speaking Tuesday at The Wall Street Journal’s Future of Everything Festival, said the speed of AI research could accelerate from its already-rapid pace.
> “The p...
Business Insider: [China is building a military capable of taking over Taiwan by 2030, US intel chief says](https://www.businessinsider.com/china-building-military-capable-take-taiwan-by-2030-avril-haines-2022-5)
> A top US intelligence official said China is set on building a military capable of taking over Taiwan by 2030.
> "It's our view that they are working hard to effectively put themselves into a position in which their military is capable of taking Taiwan," Director of National Intelligence Avril Haines told the Senate Armed Services Committee ...
Mobile service live trials with a view to practical use of level 4 technology are already underway. The central government aims to start the service in some areas around fiscal 2022, and aims to make it common nationwide by 2025.
arXiv: [A Neural Network Solves and Generates Mathematics Problems by Program Synthesis: Calculus, Differential Equations, Linear Algebra, and More](https://arxiv.org/abs/2112.15594)
On page 2:
> The MATH dataset (4) is a benchmark to measure the mathematical problem-solving ability of models and is specifically designed to offer a further challenge, in light of recent progress with word problems. This dataset is sourced from high school math competitions, such as the AMC 10, AMC 12, and AIME. State-of-the-art Transformers such as GPT-3 pre-trained only...
Washington Post: [Biden sends every signal he is running again](https://www.washingtonpost.com/politics/2022/06/17/biden-signals-run-for-reelection/)
> His advisers are studying plans for a reelection launch and flooding key states with cash. But they can’t shake worries within their party on his readiness for another campaign.
> President Biden’s advisers have been studying a spring 2023 reelection announcement that would echo the timetable of former president Barack Obama. They have flooded 2024 battleground states with millions of dollars to build u...
The Next Platform: [Intel Aims For Zettaflops By 2027, Pushes Aurora Above 2 Exaflops](https://www.nextplatform.com/2021/10/27/intel-aims-for-zettaflops-by-2027-pushes-aurora-above-2-exaflops/)
> “But to me, the other thing that’s really exciting in the space is our Zetta Initiative, where we have said we are going to be the first to zettascale by a wide margin,” Gelsinger told The Next Platform. “And we are laying out as part of the Zetta Initiative what we have to do in the processor, in the fabric, in the interconnect, and in the memory architecture ...
The Manhattan district attorney is continuing to investigate Mr. Trump, but knowledgeable people say charges are unlikely to occur in the foreseeable future, if ever.
This comment was originally posted on Donald Trump
ChatGPT is passing exams:
[ChatGPT passes the 2022 AP Computer Science A free response section](https://gist.github.com/Gaelan/cf5ae4a1e9d8d64cb0b732cf3a38e04a)
> For fun, I had ChatGPT take the free response section of the [2022 AP Computer Science A exam](https://apcentral.collegeboard.org/media/pdf/ap22-frq-computer-science-a.pdf). (The exam also has a multiple-choice section, but the College Board doesn't publish this.) It scored 32/36.
MedRxiv: [Performance of ChatGPT on USMLE: Potential for AI-Assisted Medical Education Using Large Language Model...