Simon Speech To Text
Kayleen
Alternatives? CMUsphinx is just a toolkit which simon uses, and freespeech AKA OpenMindSpeech is less. julius, another toolkit simon uses, exists and is in the 18.04 repositories, but is nowhere near as well documented as simon. Installing julius is simple: open a terminal window and type in
Charlie Holtz published an astonishing demo today, where he hooked together GPT-Vision and a text-to-speech model trained on his own voice to produce a video of Sir David Attenborough narrating his life as observed through his webcam.
I tried a couple of samples, but by far the best result I got was using the audio from my recent 10 minute talk at GitHub Universe. This isn't surprising: the Universe A/V team are extremely good at their jobs.
I already had a copy of the video downloaded from YouTube (using yt-dlp. I opened it up in QuickTime Player on macOS, trimmed it to just the ten minutes of my speaking and used File -> Export As -> Audio Only... to extract a 9m40s 8.5MB .m4a file.
I also set the description field to "40 year old British male, enthusiastic, technical". I don't think this description has any impact on the voice generation, I think it's just metadata for your own library of voices.
I hereby confirm that I have all necessary rights or consents to upload and clone these voice samples and that I will not use the platform-generated content for any illegal, fraudulent, or harmful purpose
El anuncio ms importante del OpenAI DevDay de la semana pasada (y hubo MUCHOS anuncios) fueron los GPT. Los usuarios de ChatGPT Plus ahora pueden crear sus propios bots de chat GPT personalizados con los que otros suscriptores de Plus pueden hablar.
It is great that the OECD recognises the high performance across UK government communications that I see every day. I have recently completed a series of visits to each departmental communications team and the level of dedication and expertise I saw, as well as the variety of ways in which we are seeking to innovate, makes me proud to lead GCS.
But the OECD is right to note that we have further to go to support all teams within GCS to operate at their very best. I want to briefly touch upon a few of the actions we are taking to build on the progress already made.
We are building on the recent release of the updated GCS policies and standards known as the Modern Communications Operating Model by developing a new self-assessment tool that will help all communications teams, regardless of their size and context, assess their current performance and identify areas of strength and opportunities for further development. This tool will be released in March, and will look to join up different teams across government to support one another through peer review.
On upskilling, GCS Advance is a substantial new learning and development programme that will deliver a measurable step change in the skills of UK government communicators. It will operate at apprentice, practitioner, expert and leader levels with a focus on digital and data skills, mandatory modules on AI at every level and training in agile management techniques at expert and leader levels. The practitioner level pilot is almost complete and will roll out from April, and the Expert level programme was fully subscribed in less than two weeks during January. The ambition is for 2,500 GCS members to be part of a GCS Advance programme by March of next year.
On innovation, colleagues in GCS will know that this is an area about which I am deeply passionate. Over the last two years GCS has focussed on harnessing new communications technology to drive better outcomes and improving digital and data skills.
GCS has also helped to pioneer the use of AI in government. AI tools can help communicators to get the right message, to the right audience at the right time; develop more two-way, personalised and inclusive communication; and act as a co-pilot to transform productivity and effectiveness.
The ambition is to build more GCS data and insight into the model. For example by training it on focus group and polling data to predict how different groups might react to a specific government announcement, or training it on historic media queries to predict media questions and suggest answers. This new technology could allow communications professionals to complete tasks in minutes that take hours today.
The GCS focus on innovation has already moved it from a function that was seen as lagging behind the private sector, to one which is increasingly recognised as world-leading in its use of new technology. I am committed to continuing this progress.
And the UK also scored lower than comparator countries on issues of responsiveness and integrity. Fewer than one-third (30%) of the UK population thought it was likely that a national policy would be changed if the majority of people expressed a view against it. Half (51%) of the UK population thought a change was unlikely, compared with an OECD average of 40%.
OECD research shows that trust in government is strongly associated with people feeling that they have a say in what Government does. Rebuilding trust is about more than Government delivering on what it says it will do. It is also about how it governs.
GCS has an important role to play in considering how organisational listening could be used more across government to enhance how open and participative UK government is, including by considering international examples of participative and deliberative processes such as citizens juries or assemblies. It is interesting that the Republic of Ireland, which has been at the forefront of these innovations, is close to the top of the OECD rankings for public trust in government.
New technology could support more engagement with the public in the early stages of policy development. Government consultations could be made more accessible. AI could quickly summarise long consultation documents; translate policy and receive responses in any language; summarise the main suggestions and points of concern for policy makers; and give citizens a tailored response explaining how their views were taken into account when developing policy.
As noted by the scan, I am pleased with the progress made by GCS to increase our measurement and evaluation of communications activity. Since I joined Government I have seen significant improvements in how our communications campaigns are measured and evaluated, and how the findings support further refinement and development.
Instead, communications should be seen as a critical lever for government. It should be thought of alongside legislation, regulation, taxation and spending as a tool that Ministers can use to achieve policy goals.
There are three roles for modern government communications: informing the public about the actions of the government; supporting better policy or service design through public engagement and insight; and changing behaviour for the public good or supporting operational delivery through campaigns.
The scan warns that the UK is not immune to the trend towards politicisation that the OECD has also observed in other countries. It also reminds us that the trustworthiness of public communications is essential, particularly in order to support the countering of disinformation and give the public confidence that government communications and its underpinning technologies, such as ad targeting and behavioural insights, are being used responsibly and in the public interest.
To ensure everyone in GCS has a strong understanding of the rules around propriety and ethics we have developed a new, mandatory online course for GCS members on Propriety & Ethics. The course was launched in November - and more than half of GCS members from ministerial departments have completed the training so far.
Building resistance against mis and disinformation is also a key area of focus, particularly given that there have been recent examples from other elections where instances of mis and disinformation have caused confusion. Although this is an area where we have made progress, most notably through the use of the RESIST 2 Counter Disinformation Toolkit, it is vital that we work to prepare for, and if needed respond to, instances of mis and disinformation in the run-up to or during the next election.
Finally, throughout this speech I have spoken about the benefits that innovation and greater use of new technology can bring. But, as OECD recommends, it is vital we give the public confidence that these technologies are being used responsibly and in the public interest. We are placing our ethical values at the heart of our approach to innovation, which will be outlined through our upcoming Data and Innovation Strategy.
As part of the Strategy we are developing an Ethical Decision Making Framework for Responsible Innovation, which will enable all government communicators to apply these principles when considering the use of new technology.
Cookies: We use our own and third-party cookies to improve your experience of our website. Cookies remember your preferences and track site usage. By continuing, you accept their use.
We have reached a fork in the road for audiobooks. While they have never been more popular, as people listen to a wide variety of audio content on the go and at the gym, it is now possible to use synthesized voices to make spoken-word versions of any kind of text that sound sort of real. These voices are much better than the text-to-speech voices you can already access on your computer.
Audiobooks are big business, projected to reach $35 billion by 2030. About 65% of audiobooks are fiction, and genre fiction - mystery, romance, thriller, fantasy - is by far the most widely listened to type of book. But personal memoirs, read by their authors, also attract. Think of recent books by Michelle Obama and Price Harry; there's a sense of personal connection when listening to a well-known person narrate their own book.
For authors looking to maximize their revenue streams, audiobooks are a great way to tap into readers who might not otherwise discover them. But this comes at a price. Having a professional narrator read a book costs several thousand dollars.
c80f0f1006