Scholarly publishing & Research
- Will Building LLMs Become the New Revenue Driver for Academic Publishing? This optimistic take on AI’s effect on the industry zeros in on the need for LLMs to be fed by high-quality, vetted content. “If we continue relying on general purpose AI models whose sources tend to be mysterious, we run the risk of rubbish in = rubbish out.” (The Scholarly Kitchen, August 8, 2023)
- AI is speeding up scientific discoveries and helping to spot new ideas: AI offers significant opportunities in scientific research and innovation, while presenting some serious challenges. “They're pointing mathematicians toward promising solutions, generating drug candidates that can then be tested experimentally, predicting the structures of proteins, and could help to filter massive amounts of data generated from one of the world's largest experiments.” (Axios Science, August 17, 2023)
- How should we approach the AI revolution? Introduced as ‘Captain Optimism’ for his positive view on the use of AI, Dustin Smith of Hum said he believes that tools such as large language models (LLMs) can help us to broaden and engage our audience. “AI tools have the potential to streamline our workflows, allowing us more time to focus on outstanding issues such as research integrity." (Research Information, June 27, 2023)
- Multi-Agency Research and Development Priorities for the FY 2025 Budget: The top priority in an August OSTP memo is to advance trustworthy artificial intelligence technology that protects people’s rights and safety accelerates the Nation’s progress. (OSTP, August 17, 2023)
- Are HIT-backed AI Research Integrity Solutions the Need of the Hour? Human intelligence tasks (HITs) combined with AI tools could offer "a reliable and scalable solution for maintaining research integrity within the scholarly record." (The Scholarly Kitchen, August 3, 2023)
- Forget SEO: Why ‘AI Engine Optimization’ may be the future: Introducing the idea of AI Engine Optimization and what it might mean for content creators (including publishers). (VentureBeat, August 14, 2023)
- ChatGPT-like AIs are coming to major science search engines: Like ChatGPT, Bard, and Bing, scholarly publishing databases like Scopus, Dimensions and Web of Science are introducing conversational AI search. (Nature, August 2, 2023 – paywalled)
Legal & Political
- Terms-of-service land grab: Tech firms seek private data to train AI: Absent a federal privacy law, battles are already ensuing over companies' rights to use customer data in their training materials; companies working with AI should be careful in formulating their Terms & Conditions. (Axios AI, August 18, 2023)
- AI-generated art cannot be copyrighted, rules a US federal judge: The judge added that "human authorship is a bedrock requirement of copyright." (The Verge, August 19, 2023)
- OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling's Harry Potter series: According to new research, ChatGPT has begun to avoid responding to prompts with phrases from copyrighted works, amid growing scrutiny (the irony being that this implicitly indicates the copyrighted works were used in its LLM!). (Business Insider, August 15, 2023)
- The EU Artificial Intelligence Act: Meanwhile, EU policymakers are working on finalizing the EU AI Act—"a proposed European law on artificial intelligence – the first comprehensive law on AI by a major regulator anywhere”—with goal to finalize by end of 2023 (The AI Act, August 2023)
General
- The Associated Press sets AI guidelines for journalists: The Associated Press developed guidelines for reporters and editors on using AI. “Any result from a generative AI platform “should be treated as unvetted source material” and subject to AP’s existing sourcing standards.” (The Verge, August 16, 2023)
- Will GPT models choke on their own exhaust? A new paper (summarized in this blog post) warns of “model collapse” where AIs will ingest more and more content created by AIs until the whole thing turns into gibberish. “LLMs are like fire – a useful tool, but one that pollutes the environment. How will we cope with it?” (Light Blue Touchpaper, June 6, 2023)
- How AI-generated software could turn your competitive moat into a puddle: Generative AI will dramatically speed up product and software development. A peek at what it might look like, and likely differentiators for technology providers in the future. (Mind the Product, August 3, 2023)
- Two perspectives on how transformative LLMs will be: Ian Mulvany, CTO of The British Medical Journal writes on his blog about some of the trends he predicts will play out with large language models. (Ian Mulvany, August 15, 2023)
To keep up on future reading lists and other community & industry updates, subscribe to our monthly email newsletter.