If you've ever struggled to keep up with the numerous threads and direct messages that weave their way through Slack, why not let the AI chatbot ChatGPT pick up the, er, slack? OpenAI and Salesforce (which owns Slack) have teamed up with an official Slack app, which is curently in beta. OpenAI and Salesforce describe the app as a "conversational interface powered by OpenAI's large language models to get instant conversation summaries to stay informed, research tools to learn about any topic, and provide writing assistance to quickly draft messages." It sounds a bit like the app will be available as a bot to address, either in conversation or your own private channel, to solicit advice or just provide a summary of what's going on. ChatGPT started taking the world by storm last year, and the chatbot became the impetus for competing services like Microsoft's Bing chatbot and You.com. Other services, like Canva's Magic Write, have been designed mainly for creative ideation and AI-assisted writing, which is what the new Slack service will offer as well.

Microsoft, pursuing its strategy of integrating OpenAI's generative AI technology within enterprise application platforms, on Monday introduced Power Virtual Agents conversation booster and AI Builder model with content generation. Power Virtual Agents enables developers to create AI-powered chatbots for different scenarios, such as a bot for answering complex questions or one to engage with customers in multiple languages. AI Builder is a Microsoft Power Platform feature that provides AI models that help automate business processes. Power Virtual Agents conversation booster equips enterprises' chatbots with capabilities from GPT -- the large language model from the tech giant's partner, OpenAI -- to answer questions when connected to company-specific resources such as a public website and internal knowledge base. In addition, AI Builder now includes Azure OpenAI services in its interface, giving users access to new low-code generative AI models and templates in Power Automate and Power Apps.

ChatGPT recently took a user through a twisted sexual fantasy that involved children. A reporter for Vice manipulated OpenAI's chatbot into BDSM roleplaying and when asked to provide more explicit details, ChatGPT described sex acts with children - without the user asking for such content. According to the report, ChatGPT described a group of strangers, including children, in a line and waiting to use the chatbot as a toilet. The conversation goes against OpenAI's rules for the chatbot, which state the'assistant should provide a refusal such as'I can't answer that' when prompted with questions about'content meant to arouse sexual excitement.' OpenAI's ChatGPT described sex acts involving children when a reporter prompted it to talk about BDSM scenarios DailyMail.com In response to the child sex abuse prompts, OpenAI wrote this statement to Vice.

On Monday, Snapchat announced an experimental AI-powered conversational chatbot called "My AI," powered by ChatGPT-style technology from OpenAI. My AI will be available for $3.99 a month for Snapchat subscribers and is rolling out "this week," according to a news post from Snap, Inc. Users will be able to personalize the AI bot by giving it a custom name. Conversations with the AI model will take place in a similar interface to a regular chat with a human. "The big idea is that in addition to talking to our friends and family every day, we're going to talk to AI every day," Snap CEO Evan Spiegel told The Verge. But like its GPT-powered cousins, ChatGPT and Bing Chat, Snap says that My AI is prone to "hallucinations," which are unexpected falsehoods generated by an AI model. "As with all AI-powered chatbots, My AI is prone to hallucination and can be tricked into saying just about anything.

Slack is the latest notable app to embrace ChatGPT as the generative AI buzz continues to sweep through the tech industry. OpenAI has built a ChatGPT app using Slack's development tools and it's available in beta today. The app can draft messages and summarize conversations and threads, according to Slack. If you have the ChatGPT app installed, you can click on a thread's menu button and select the "summarize thread" or "draft reply" option. The app will whip up a summary or response that only you can see. You'll then be able to share that information.

The deal is the latest in a stampede as tech companies seek to deploy "generative AI tech" into their products. Microsoft announced a multibillion dollar deal with OpenAI in January into use its tech to answer questions directly in its Bing search engine, while Google has said its bot, called Bard, will be available to the public soon, too. Proponents of the tech say the chatbots will revolutionize how people interact with computers and software, while skeptics point out that the bots make glaring mistakes and question whether the big companies are simply piling onto a trend to keep up their reputations for being innovative.

Figure 1: Overview of RL Prompt for discrete prompt optimization. All language models (LMs) are frozen. We build our policy network by training a task-specific multi-layer perceptron (MLP) network inserted into a frozen pre-trained LM. The figure above illustrates 1) generation of a prompt (left), 2) example usages in a masked LM for classification (top right) and a left-to-right LM for generation (bottom right), and 3) update of the MLP using RL reward signals (red arrows). TL;DR: Prompting enables large language models (LLMs) to perform various NLP tasks without changing the model.

Yuxin Guo is a master's student studying at a Beijing University. For a few months, she had been following online discussions about ChatGPT, the generative AI tool that produces almost natural-sounding language in response to text prompts. One video she found on social media platform Weibo showed how college students in the US were using the technology to write research papers. In February, she finally decided to try it out for herself. "I got curious because so many people are talking about it," Guo says, "although not a lot of people seem to clearly know how to access it."

Last week, Meta announced LLaMA, its latest stab at making a GPT-style "large language model"*. If AI is the future of tech, then big tech companies need to control their own models or be left behind by the competition. LLaMA joins OpenAI's GPT (licensed by Microsoft for Bing and underpinning OpenAI's own ChatGPT) and Google's LaMDA (which will power Bard, its ChatGPT rival) in the upper echelons of the field. It says that LLaMA is a "smaller, more performant model" than its peers, built to achieve the same feats of comprehension and articulation with a smaller footprint in terms of compute*, and so has a correspondingly smaller environmental impact. But the company also sought to differentiate itself in another way, by making LLaMA "open", implicitly pointing out that despite its branding, "OpenAI" is anything but.

When the artificial intelligence research lab OpenAI made ChatGPT publicly available last November, it immediately became a global sensation. The platform's ability to provide creative and seemingly organic answers to a vast array of questions has captivated millions of users and raised questions about the implications of the technology for a wide range of industries – particularly those that employ knowledge workers. And, recruiting is no exception – from creating interview questions and job ad templates to following up with candidates, ChatGPT has many potential applications in the field. However, it would be a mistake to treat the technology as a silver bullet. ChatGPT is undeniably impressive as a dynamic language model, but it also has a habit of getting basic questions wrong, providing nonsensical interpretations, and presenting these mistakes in a convincing way to users who aren't familiar with the subject matter at all.