Google Cloud London<\/a> this month.<\/p>\n\u201cWe think it’s [saved us] over 20 person years of audio transcription…in a matter of weeks.\u201d<\/p>\n
Although a person checks each recording, the AI transcription\u2019s accuracy rate of 99% is \u201csignificantly better than a human,\u201d Nick notes.<\/p>\n
\u201cIt really changes how you conduct research\u201d<\/p>\n
A transcript is a simple text-based version of a recording. You can search it for basic information – \u201cthe names of ships or the names of people or particular dates or a particular battle\u201d – but it\u2019s very limited as a research tool.<\/p>\n
Adding Google’s Gemini AI has changed the game. Now, historians can dig much more deeply.<\/p>\n
\u201cYou can use an LLM to say, \u2018Was this person scared at any point? Do they have any funny stories?’… ‘Why might that person have been scared?’ or \u2018What happened next after the funny story?\u2019<\/p>\n
\u201cIt really kind of changes how you conduct research…[and] one of the biggest takeaways is it\u2019s transformed our thinking about how people might access collections and museum collections in the future.\u201d<\/p>\n
So far the system has transcribed more than 8,200 interviews across nearly 46,000 separate recordings and 15 languages. Navigating those, even with AI\u2019s help, is no small task.<\/p>\n
That\u2019s where tags, or entities to use their official term, come in. The LLM assigns these when it recognises a key point in a recording, like \u201cWorld War 2,\u201d \u201cD-Day\u201d or \u201cPurley,\u201d to make it easier to find related artefacts.<\/p>\n
The IWM has its own list of about 400,000 entities, but many are synonyms (World War 1, First World War). Running an AI over that would reduce the list by \u201cabout 50 or 60%\u201d.<\/p>\n
\u201cA surprising amount of our search traffic comes from people Googling World War 11,\u201d Nick notes.<\/p>\n
Reimagining research sounds great, and interactive transcripts could completely change how the public uses the IWM\u2019s digital archive, but there is one issue: cost. As a non-profit, the museum has limited resources but a large following: its website records 20 million visitors every year.<\/p>\n
\u201cWhen we make a product accessible, often, a lot of people come and use it straight away – and with an LLM, you’re paying for every for every prompt… In retrospect, building it was probably the cheapest thing [we\u2019ll do].\u201d<\/p>\n
Talks are ongoing with Google to find a solution, but as a stopgap the IWM is being flexible with the models it uses: Gemini 1.5 if Gemini 2 is \u201coverkill\u201d, 2.5 or 2.5 Pro if 2 \u201cisn\u2019t cutting it.\u201d<\/p>\n
For now, balancing cost with ambition remains a challenge. But what good is a world-class digital archive if it\u2019s not being used? AI might not be a perfect answer, but it\u2019s a promising tool in making sure these voices are heard again.<\/p>\n","protected":false},"excerpt":{"rendered":"\u2018World War 11\u2019 and other reasons the IWM needed AI Image: AI has saved the IWM \u201c20 person…\n","protected":false},"author":2,"featured_media":42310,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[46],"tags":[181,33966,191,2788,168,4430,33967,33968,8184,33969,74,33970],"class_list":{"0":"post-42309","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-computing","8":"tag-artificial-intelligence","9":"tag-capgemini","10":"tag-computing","11":"tag-data","12":"tag-google","13":"tag-google-cloud","14":"tag-imperial-war-museums","15":"tag-iwm","16":"tag-llm","17":"tag-nick-hodder","18":"tag-technology","19":"tag-transcript"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/42309","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/comments?post=42309"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/42309\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media\/42310"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media?parent=42309"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/categories?post=42309"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/tags?post=42309"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\".\/media_1704ddaf5708232d72ad942aac8ed0022f6a0534a.jpg?width=750&format=jpg&optimize=medium\")
<\/p>\n