Google has developed a new AI tool to help marine biologists better understand coral reef ecosystems and their health, which can aid in conversation efforts.
The project began by inviting the public to listen to reef sounds via the web.
By crowdsourcing this activity, Google was able to create a library of new fish sounds that were used to fine-tune the AI tool, SurfPerch.
Although bird sounds and reef recordings are very different, there were common patterns between bird songs and fish sounds that the model was able to learn from, they found.
The project continues today, as new audio is added to the Calling in Our Corals website, which will help to further train the AI model, Google says.
It’s been 20 years since Shira Yevin, the lead singer of punk band Shiragirl drove a pink RV into the Vans Warped Tour grounds, the now-defunct punk rock festival notorious for being male-dominated.
After Shiragirl Stage, Yevin went on a “lifelong crusade for carving out a space for women in music and other male-dominated industries,” she told TechCrunch.
This began with the launch of Gritty in Pink in 2020, a Live Nation-backed platform designed to empower women in the music industry.
Like freelance platforms Fiverr and Upwork, users can publish listings to highlight their services on InPink’s marketplace and connect with major brands.
Yevin said InPink is already working with one sports client, Angel City Football Club (the LA-based women’s soccer league), to help them hire talent for pre-shows.
Not all generative AI models are created equal, particularly when it comes to how they treat polarizing subject matter.
They found that the models tended to answer questions inconsistently, which reflects biases embedded in the data used to train the models, they say.
“Our research shows significant variation in the values conveyed by model responses, depending on culture and language.”Text-analyzing models, like all generative AI models, are statistical probability machines.
Instrumental to an AI model’s training data are annotations, or labels that enable the model to associate specific concepts with specific data (e.g.
Other studies have examined the deeply ingrained political, racial, ethnic, gender and ableist biases in generative AI models — many of which cut across languages, countries and dialects.
Greptile, an early stage startup from a group of recent Georgia Tech grads, decided to take a different approach: using AI to help developers understand the code base.
Greptile CEO and co-founder Daksh Gupta says the Greptile bot is like having a highly experienced coworker who has a deep understanding of your code.
“So we’re building AI tools that understand large code bases at companies because as time goes on, and multiple programmers work on the codebase, it tends to get very difficult to understand,” Gupta told TechCrunch.
Once the repositories have been indexed by the system, you add a natural language query such as, how does the authentication work in this code base,” he said.
The startup launched last July after the founders came up with the idea for the company at a hackathon.
Google on Thursday said it is rolling out NotebookLM, its AI-powered note-taking assistant, to over 200 new countries, nearly six months after opening its access in the U.S.
The list of countries that NotebookLM now supports includes Australia, Brazil, Canada, India, and the U.K., as well as 208 other countries and territories.
It uses AI to help generate summaries and answer questions from documents, transcripts, notes and other sources that users can upload.
Some early users of NotebookLM in the U.S. anticipated it would support traditional note-taking apps, including Evernote and Google Keep.
Gemini 1.5 Pro also lets NotebookLM have up to 50 sources in each notebook, with 500,000 words per source.
WWDC is for developers, and much of the focus will be on iOS 18.
As the big event nears, all sorts of leaks have emerged about what iOS 18 and its rumored AI-powered apps and features have in store.
AI recaps of missed texts will also be available.
Users will also be able to change the color of app icons, perhaps as Apple’s answer to Google’s Material You.
Maps with route creationMaps will support custom route creation in iOS 18, which lets users design their own routes for a trip.
Apple will kick off its weeklong Worldwide Developers Conference (WWDC 2024) event with the customary keynote at 10 a.m. ET/7 a.m. PT on June 10.
The presentation will focus on the company’s software offerings and the developers that power them, including the latest versions of iOS, iPadOS, macOS, tvOS, visionOS and watchOS.
There’s a stream on YouTube as well, but that has a tendency to lag a bit.
And you can bet that many of iOS 18’s AI features will make their way to macOS 15, as well.
TechCrunch will be reporting on the ground at Apple Park, bringing you the news as it happens.
French AI startup Mistral is introducing new AI model customization options, including paid plans, to let developers and enterprises fine-tune its generative models for particular use cases.
Mistral has released a software development kit (SDK), Mistral-Finetune, for fine-tuning its models on workstations, servers and small datacenter nodes.
For developers and companies that prefer a more managed solution, there’s Mistral’s newly launched fine-tuning services available through the company’s API.
Compatible with two of Mistral’s models for now, Mistral Small and the aforementioned Mistral 7B, Mistral says that the fine-tuning services will gain support for more of its models in the coming weeks.
Lastly, Mistral is debuting custom training services, currently only available to select customers, to fine-tune any Mistral model for an organization’s apps using their data.
Stability AI, the startup behind the AI-powered art generator Stable Diffusion, has released an open AI model for generating sounds and songs that it claims was trained exclusively on royalty-free recordings.
Called Stable Audio Open, the generative model takes a text description (e.g.
Stability AI says that it’s not optimized for this, and suggests that users looking for those capabilities opt for the company’s premium Stable Audio service.
Stable Audio Open also can’t be used commercially; its terms of service prohibit it.
And it doesn’t perform equally well across musical styles and cultures or with descriptions in languages other than English — biases Stability AI blames on the training data.
And like all productivity tools, the ClickUp team has also heard the siren song of artificial intelligence.
The company has now launched what it calls “ClickUp Knowledge Management,” which combines a new wiki-like editor and with a new AI system that can also bring in data from Google Drive, Dropbox, Confluence, Figma and other sources.
With that, the company aims to build a tool that can rival other popular services like Notion and Atlassian’s Confluence.
The result, ClickUp argues, is a system that brings together the best of Notion, Confluence and Glean to allow users to quickly create documents.
This now enables the ClickUp Knowledge Management to perform retrieval augmented generation (RAG) — which has quickly become the industry standard for augmenting large language models (LLMs) with additional and up-to-date information.
