Today, TechHive is launching Smart Answers, a chatbot tool that helps you get more from our content. It’s built using Generative AI and content written by our human editors.
The way we interact with content is changing. It wasn’t so long ago you would have sifted through a printed magazine for advice on the latest consumer technology, yet it felt like a revolution when those old mags switched over to digital and online editions. These days, everything you could ever want to read is on the internet—or more likely on YouTube or TikTok.
The difficulty comes in finding trustworthy, up-to-date information that precisely and immediately answers your question, and most of us tend to rely on search engines and algorithms to figure that out for us, with varying degrees of success. But faced with a page full of links to visually similar content, you’re still a step away from the information you need—or two steps, in fact, if you need to refine your query.
Smart Answers is our new reader service that turns content discovery on its head. You control the questions, the answers, and the delivery. It’s like having a TechHive editor at your beck and call, allowing you to request and receive specific content on demand. So, why wait for the content to find you?
We’ve worked in partnership with MISO.AI to develop Smart Answers, a GenAI tool (see What is GenAI?) that draws only on Foundry‘s complete catalog of English-language content (written by humans) to answer your natural-language questions. You’ll not only benefit from the expert knowledge of TechHive editors, but those of PCWorld, Tech Advisor and Macworld too, collectively covering the entire consumer technology sphere.
Smart Answers is fast and efficient, which means you don’t have to scroll through pages of information to find what you need. You can ask it anything, from what’s the best solution for video streaming to which doorbell to buy for your smart home. You’ll get an answer in seconds.
Ask it a follow-up question too, if you like, or browse the recommended reading for more detail on the topic. Smart Answers is clever enough to work out what you may want to ask next, and it will even show you best pricing for related products and services to avoid you having to search it out yourself.
Smart Answers has been trained on the type of tech-related questions our readers are likely to ask. Ask it a silly or irrelevant question, and you’ll either get a silly or irrelevant answer, or no answer at all. Our priority is to ensure Smart Answers is able to adequately respond to questions on the topics we cover.
Though Smart Answers is intrinsically linked to our editorial content, our editorial function stands independently. No article published on this site has been created using GenAI.
Our editors have conducted extensive testing over many months, with their feedback used to retrain the model. We wouldn’t be rolling this out today if we were not confident that the answers it provides are as trustworthy as if they had come directly from our editorial team.
This cycle of testing, feedback and retraining is ongoing, and will be important as we continue to add functionality to Smart Answers. Today our Smart Answers page goes live, but over the coming week you will also find a Smart Answers block on our home page and within our articles. And this is just the start of a process that will ultimately put you in charge of your own content experience across Foundry-owned sites.
Of course the real testing comes from you. We’ll be analyzing all search data collected by Smart Answers, and cross-checking the quality of responses, but we’d also love to hear your feedback. You can get in touch with our Editorial Director (EMEA & LatAm) Marie Black or Director of Product and Data Neil Bennett over email.
How does Smart Answers work?
Smart Answers is based on a GenAI platform, built by our partner MISO.AI. Unlike ChatGPT, Google Bard and other GenAI tools that source their information from all over the web, Smart Answers has been trained only on content written by Foundry journalists. It responds to queries posed in natural language with a summarized answer and links to related information.
What is GenAI?
Short for Generative AI, GenAI is a type of artificial intelligence that is able to mimic the neural networks of the human brain and, through machine learning of existing content, generate entirely new content in the form of text, images, video and audio. When given a prompt, such as a piece of natural-language text, the computational power of GenAI allows it to deliver an answer much faster than can the human brain.
Can I trust Smart Answers?
Yes. Smart Answers provides answers based on articles written by Foundry journalists and does not use data from any external sources, which could be unreliable or subject to copyright concerns. Extensive testing has gone on behind the scenes to help train the model, so we are satisfied that it is able to provide reliable information. We’ve also worked hard with partner MISO.AI to eliminate “hallucinations.” which are misleading or wholly incorrect answers that may be given when the answer is not known or the data required to answer the question is incomplete. If Smart Answers is unable to answer your question, it will tell you so.
Is Smart Answers a threat to our editorial team?
Absolutely not. Smart Answers cannot exist without the expert content written by Foundry journalists. Its ability to stay useful and relevant is entirely dependent on it being able to learn from their content, and thus it is not the first step in some evil plan to replace our team of editors and cut costs. Smart Answers is a supplementary service for our readers, designed to aid content discovery and enhance the user experience. Though we will make revenue from any ads displayed on the page, and earn commission on links to purchase some products and services, Smart Answers also costs us money to develop and run.
Is Smart Answers being used to create editorial content on TechHive?
No. Editorial articles feed into Smart Answers, but Smart Answers is not used for the creation of editorial articles. But in the future we may use query data to come up with ideas for articles and FAQs
What data is Smart Answers collecting?