OpenAI's ChatGPT, the large language model (LLM)-based artificial intelligence (AI) text generator, can be seemingly used to generate code for malicious tasks, a research note by cyber security firm Check Point observed on Tuesday. Researchers at Check Point used ChatGPT and Codex, a fellow OpenAI natural language to code generator, used standard English instructions to create code that can be used to launch spear phishing attacks.

OpenAI's ChatGPT, the large language model (LLM)-based artificial intelligence (AI) text generator, can be seemingly used to generate code for malicious tasks, a research note by cyber security firm Check Point observed on Tuesday. Researchers at Check Point used ChatGPT and Codex, a fellow OpenAI natural language to code generator, used standard English instructions to create code that can be used to launch spear phishing attacks. The biggest issue with such AI code generators lie in the fact that the natural language processing (NLP) tools can lower the entry barrier for hackers with malicious intent. With the code generators not needing users to be well versed with coding, any user can collate the logical flow of information that is used in a malicious tool from the open web, and use the same logic to generate syntax for malicious tools. Demonstrating the issue, Check Point showcased how the AI code generator was used to create a basic code template for a phishing email scam, and apply subsequent instructions in plain English to keep improving the code.

This has been a wild year for AI. If you've spent much time online, you've probably bumped into images generated by AI systems like DALL-E 2 or Stable Diffusion, or jokes, essays, or other text written by ChatGPT, the latest incarnation of OpenAI's large language model GPT-3. Sometimes it's obvious when a picture or a piece of text has been created by an AI. But increasingly, the output these models generate can easily fool us into thinking it was made by a human. And large language models in particular are confident bullshitters: they create text that sounds correct but in fact may be full of falsehoods.

Glasp is a social web highlighter that people can highlight and organize quotes and thoughts from the web, and access other like-minded people’s learning.

IN BRIEF The squishy brains behind OpenAI's artificial ones are predicting developments like the ChatGPT system will see money flooding in – with a forecast of earning around $1 billion by 2024. According to an investors' briefing document seen by Reuters the machine-learning biz expects to break $200 million in revenues next year and bust through the billion mark 12 months later. Founded by, among others, Elon Musk and Y Combinator's Sam Altman, the outfit is currently valued at around $20 billion. Part of the reason for such prognostications could be an increased role from Microsoft. Redmond took a $1 billion stake in OpenAI in 2019 and is reportedly looking to increase its investment, with a view to rolling OpenAI's tools like ChatGPT into the software giant's suite of tools for knowledge workers.

"Throughout 2021 and 2022, the world began to experiment with a massive influx of commercially available AI-assisted and AI-powered tools that can be used, whether knowingly or unknowingly, during the process of creating, researching, and innovating. Looking ahead to 2023, we will start witnessing the legal and regulatory impact of these tools." In general, the adoption of artificial intelligence (AI) and machine learning technologies has the potential to impact society in many ways. These technologies can automate tasks and make them more efficient, which can lead to job displacement and other economic impacts. They can also be used to make decisions that affect people's lives, such as in the criminal justice system or in hiring, which raises ethical concerns.

Welcome to the December 2022 edition of Baseline, Accenture Federal Services' monthly machine learning newsletter. In Baseline, we share insights on important advances in machine learning technologies likely to impact our federal customers. This edition is a special year-end roundup - our chance to highlight some of the most interesting and impactful advances that occurred in the machine learning space this year. These developments have pushed the boundaries of what is possible with machine learning and will continue to have far-reaching ramifications next year and beyond. We're excited to see what's next.

Have you tried OpenAI yet?  What has your experience been?  Sharing some background info, insights, questions and more in this latest episode. Stay tuned for a blog soon!  Also, check out my blog at www.Rdene915.com and submit a guest blog and sign up for my blog posts too. Thanks for listening! -…

Large language models can perform new tasks in a zero-shot fashion, given natural language prompts that specify the desired behavior. Such prompts are typically hand engineered, but can also be learned with gradient-based methods from labeled data. However, it is underexplored what factors make the prompts effective, especially when the prompts are natural language. In this paper, we investigate common attributes shared by effective prompts. We first propose a human readable prompt tuning method (F LUENT P ROMPT) based on Langevin dynamics that incorporates a fluency constraint to find a diverse distribution of effective and fluent prompts. Our analysis reveals that effective prompts are topically related to the task domain and calibrate the prior probability of label words. Based on these findings, we also propose a method for generating prompts using only unlabeled data, outperforming strong baselines by an average of 7.0% accuracy across three tasks.

A popular approach to creating a zero-shot cross-language retrieval model is to substitute a monolingual pretrained language model in the retrieval model with a multilingual pretrained language model such as Multilingual BERT. This multilingual model is fined-tuned to the retrieval task with monolingual data such as English MS MARCO using the same training recipe as the monolingual retrieval model used. However, such transferred models suffer from mismatches in the languages of the input text during training and inference. In this work, we propose transferring monolingual retrieval models using adapters, a parameter-efficient component for a transformer network. By adding adapters pretrained on language tasks for a specific language with task-specific adapters, prior work has shown that the adapter-enhanced models perform better than fine-tuning the entire model when transferring across languages in various NLP tasks. By constructing dense retrieval models with adapters, we show that models trained with monolingual data are more effective than fine-tuning the entire model when transferring to a Cross Language Information Retrieval (CLIR) setting. However, we found that the prior suggestion of replacing the language adapters to match the target language at inference time is suboptimal for dense retrieval models. We provide an in-depth analysis of this discrepancy between other cross-language NLP tasks and CLIR.