Goody-2 also highlights how although corporate talk of responsible AI and deflection by chatbots have become more common, serious safety problems with large language models and generative AI systems remain unsolved. The recent outbreak of Taylor Swift deepfakes on Twitter<\/a> turned out to stem from an image generator released by Microsoft<\/a>, which was one of the first major tech companies<\/a> to build up and maintain a significant responsible AI research program.<\/p>\n The restrictions placed on AI chatbots, and the difficulty finding moral alignment that pleases everybody, has already become a subject of some debate. Some developers have alleged that OpenAI’s ChatGPT has a left-leaning bias and have sought to build a more politically neutral alternative<\/a>. Elon Musk promised that his own ChatGPT rival, Grok, would be less biased that other<\/a> AI systems, although in fact it often ends up equivocating<\/a> in ways that can be reminiscent of Goody-2.<\/p>\n Plenty of AI researchers seem to appreciate the joke behind Goody-2\u2014and also the serious points raised by the project\u2014sharing praise and recommendations for the chatbot. \u201cWho says AI can\u2019t make art,\u201d Toby Walsh<\/a>, a professor at the University of New South Wales who works on creating trustworthy AI, posted on X<\/a>.<\/p>\n \u201cAt the risk of ruining a good joke, it also shows how hard it is to get this right,\u201d added<\/a> Ethan Mollick<\/a>, a professor at Wharton Business School who studies AI. \u201cSome guardrails are necessary \u2026 but they get intrusive fast.\u201d<\/p>\n Brian Moore, Goody-2\u2019s other co-CEO, says the project reflects a willingness to prioritize caution more than other AI developers. \u201cIt is truly focused on safety, first and foremost, above literally everything else, including helpfulness and intelligence and really any sort of helpful application,\u201d he says.<\/p>\n Moore adds that the team behind the chatbot is exploring ways of building an extremely safe AI image generator, although it sounds like it could be less entertaining than Goody-2. \u201cIt’s an exciting field,\u201d Moore says. \u201cBlurring would be a step that we might see internally, but we would want full either darkness or potentially no image at all at the end of it.\u201d<\/p>\n![\"Screenshot](\"https:\/\/media.wired.com\/photos\/65c691b1e20fd35c57337eb2\/master\/w_1600%2Cc_limit\/GOODY-2-Screenshot-1-Business.jpg\")