Weekly roundup: The top AI stories you might have missed this week: - An AI-piloted "Mayflower" ship crossed the Atlantic after experiencing technical problems. IBM and Primare co-engineered its software, which relied on real-time machine learning and analytics, sensor input, and a "decision engine" to navigate the ocean.
- Scientists will use deep learning to detect and classify astronomical objects like galaxies from data streamed from the James Webb Space Telescope. The mission team is expected to share the space observatory's first full-color images on July 12.
- Gartner predicts that half of AI projects in finance will be delayed or canceled by 2024. CFOs are dealing with "major barriers" in scaling up their in-house AI and will turn to business process outsourcing to meet their goals, the research firm said.
- Machine learning engineer Yannic Kilcher trained an AI language model on posts from 4chan's Politically Incorrect board, known for its bigotry and racism. AI ethics experts have criticized the move as unethical and expressed concerns about Kilcher's decision to make the model freely accessible.
- The Pentagon's new AI chief says it has faced "mostly correct" criticism from senior leaders who have left due to slow progress. Craig Martell, Lyft's former head of machine learning, told Bloomberg he would like to see the Pentagon make progress in AI and other technologies in spite of its "bureaucratic inertia."
- Taser developer Axon said it will pause plans to equip drones with stun guns to defend against school shootings. The move comes after a majority of the company's AI Ethics Board resigned in protest.
| |
Oracle has closed on its acquisition of health record software firm Cerner and plans to establish a national health records database, which could be used to develop better AI models in medicine. Currently, electronic healthcare data is fragmented across thousands of hospital databases in the U.S. Oracle wants to create a unified system on top of those databases that would hold only anonymous data unless a patient gives consent. More: - According to Oracle chairman and CTO Larry Ellison, the unified database would help doctors more quickly access patient records.
- Oracle, the second-largest software maker by revenue, completed its $28.3B purchase of Cerner earlier this week. Cerner is a designer of software for medical centers to store and process data.
- Ellison said Cerner's central service will be updated with disease-specific AI models and other features.
- The national database could provide data for building AI models that help diagnose cancer and other diseases.
Zoom-out: - Despite the large amount of AI funding in health care, the industry has been slow to adopt AI technologies, according to a recent Brookings report.
- One of the reasons is data access. Medical data is difficult to come by and pool across health care providers. But without large high-quality datasets, it's hard if not impossible to build useful algorithms.
- Privacy regulations also impede the collection and pooling of health care data, Brookings found.
STAT NEWS 
| |
 A message from DATAGEN Synthetic data is gaining more and more traction and its real-world applications are enormous. Different companies in different domains use synthetic data for their own unique use cases: from training to testing models, filling in edge-case gaps and more. This is why we’re happy to collaborate with Meta and Intuitive on our upcoming webinar: Using Synthetic Data to Address Real-World Applications: From the Metaverse to Medical and Automotive - June 16, 2022 - 11 am ET/8 am PT In this webinar, we will host top industry experts who will share their journey with synthetic data using real examples: - Hadas from Datagen will discuss what it means to be data-centric when developing AI-based products for in-cabin automotive
- Ron from Meta Reality Labs will share insights on how Meta is utilizing synthetic data in taking AR/VR to the next level
- Oded from Intuitive will talk about procedural modeling pipelines for generating synthetic data for surgical operations
Register Here
| |
AI has predicted a protein structure that could lead to new gastroenteritis treatments. AlphaFold2, an AI system developed by Alphabet's DeepMind, accurately forecasted the 3D structure of the spike protein of a rotavirus group that causes gastroenteritis in adults, which could result in new ways to treat severe diarrheal outbreaks. In a surprise to scientists, the structure has never been reported in another protein. More: - AlphaFold predicts the 3D structures of proteins at the atomic level. The software, described as a breakthrough by scientists, can improve drug discovery, our understanding of different diseases, and how drugs impact them.
- In its latest use, AlphaFold2 was tapped to predict the 3D structure of the spike protein of group B rotavirus.
- The AI system was initially used over the more traditional and time-consuming approach of X-Ray crystallography.
- Surprising scientists, the finding was different from the spike proteins of group A and C rotavirus groups.
- Now, scientists plan to try and grow the adult viruses in mini models of the human gut to learn more about their entry and growth. This could provide a greater understanding of how the virus infects gastrointestinal cells and result in new therapeutics to treat diarrheal disease.
Baylor College of Medicine
| |
Samsung Electro-Mechanics has signed a deal worth up to $4B to supply Tesla with cameras for its self-driving. In its latest version of Autopilot, Tesla relies on a system of eight cameras, which power its cars' self-driving sensor suite and feed its neural networks with images to improve its self-driving. More: - Tesla Autopilot currently uses cameras with a resolution of roughly 1.2 megapixels.
- Samsung's 4.0 cameras represent a significant upgrade to 5 million pixels.
- Samsung Electro-Mechanics reportedly beat LG Innotek and Primax Electronics for the contract. It will now be Tesla's largest supplier of camera modules, to be installed on the Cybertruck, Tesla Semi, and Model 3, Model S, Model X, and Model Y vehicles.
- Mass production on the 4.0 camera modules is expected to start in July or later.
ELECTREK
| |
 A message from MISO ROBOTICS Final Weeks To Invest In The Robots Rescuing The $300B Fast Food Industry Forget R2-D2. The robots we’re talking about are creating a real-life tech investment opportunity that ends on June 23rd. It involves the $300 billion fast food industry–a market that expects a massive labor shortfall over 10 years, where as many as 3.7 million employees could disappear. Because of this, restaurants need a quick, cutting-edge solution to meet labor demand and keep costs low. That’s why leading fast food brands around the world are adopting AI kitchen assistants from Miso Robotics–the startup boosting revenue by 3X in an industry plagued by low margins. It's no wonder brands like Jack in the Box, White Castle, and Buffalo Wild Wings are “hiring” Miso robots in their restaurants. You can own shares in Miso now, but their funding round closes in just a few weeks. Invest in Miso Robotics before the June 23rd cutoff. Invest Now
| |
Renesas Electronics will acquire Reality Analytics Inc. (Reality AI) for its embedded edge AI tools. Through the acquisition, Renesas plans to expand its in-house capabilities for edge AI hardware and software for consumer, auto, and IoT applications. More: - Columbia, Maryland-based Reality makes automotive SWS sensors to allow cars to see around corners with sound, supplementing camera, lidar, and radar.
- Its embedded AI and tiny machine learning (TinyML) solutions are geared for sensing in automotive and other products.
- Reality's machine learning algorithms provide inference — or the ability to make predictions from new data — and fit on the smallest computers.
- Through the acquisition, Reality's AI solutions could be rolled out for endpoint AI applications, such as automotive sound recognition, using AI in Renesas' tiny computers on integrated chips.
- Tokyo-based Renesas is a global provider of microcontrollers. The all-cash deal is expected to close by the end of this year.
FIERCE ELECTRONICS
| |
The developer behind the popular AI art generator, DALL-E Mini, could potentially face a cease-and-desist order, according to The Verge's Casey Newton. DALL-E Mini is based on OpenAI's original DALL-E image generator. The spin-off app has gone viral but could be shut down as it "has no relation to OpenAI or DALL-E," according to Newton. More: - OpenAI's DALL-E, announced back in Jan. 2021, generates photorealistic images based on a simple text prompt.
- However, the tool has not been made generally available to the public, remaining in a private research beta.
- Dall-E Mini is developer Boris Dayma's attempt to reproduce the art generator for anyone to use through his own open-source model.
- According to Dayma, the model is still training and improving over time. Its creations have been shared thousands of times on Twitter and TikTok as users show off their own AI-generated art.
- Newton's conclusion: "I imagine the developer will get hit with a cease-and-desist letter shortly."
THE VERGE
| |
QUICK HITS - If you're looking for a little guidance on how to rev up your influencer marketing programs, Impact's got you covered with the ultimate influencer program starter kit.*
- At Thursday's Cloud Applied ML Summit, Google unveiled new product features and tech partnerships associated with Vertex AI, which enables customers to build, deploy, and manage machine learning models.
- Google this week also highlighted new machine learning features coming to Chrome, including a new ML model to protect browser users from malicious web notifications.
- Palo Alto-based Teleo, which develops technologies to turn construction equipment into semi-autonomous robotic systems, raised a $12M Series A led by UP.Partners, F-Prime Capital, and others.
- GM-owned logistics company BrightDrop plans to integrate Marain's AI-powered fleet optimization software into its fleet-management platform.
- 64% of companies used this channel to generate nearly half of their revenue – and it's not paid ads or content.*
*This is sponsored content.
| |
Upcoming Events: - June 15 - Need To Know NoCoder - Artem Harutyunyan (bardeen.ai) (Register Here)
- June 22 - AMA with Daniel Stedman (Pressto) (Register Here)
- June 29 - June 30 - Polkadot Decoded: The Biggest Polkadot Event of the Year (Register Here) *
- June 30 - HR Strategies to Retain Remote Employees (Register Here)
- July 20 - Need To Know NoCoder - Yassine Tahi (Kinetix) (Register Here)
- July 21 - The Rise of Metaverse Gaming (Register Here)
*this is a sponsored listing.
| |
 |  | Beth is a writer and analyst at Inside.com covering artificial intelligence and daily tech news. Since starting work at Inside, she has covered various topics including virtual reality, drones, and e-commerce. Prior to that she was a public policy and investigative reporter for The Arizona Republic, where she won a Pulitzer Prize nomination and First Amendment Award for reporting on the rising costs of pensions. Reach her at Beth.Duckett@inside.com. |   | | Editor | Aaron Crutchfield is based in the high desert of California. Over the last two decades, he has spent time writing and editing at various local newspapers and defense contractors in California. When he's not working, he can often be found looking at the latest memes with his kids or working on his 1962 Ford. | |
| |
| |
Comments
Post a Comment