AI Weekly Digest Issue #11

Welcome back to AI Weekly Digest #11, your go-to resource for the most important artificial intelligence news and events from Week 26 of 2023.

As an AI enthusiast, staying informed about the rapid advancements in the field is crucial for both personal and professional growth. Our mission is to bring you a concise and comprehensive roundup of the latest breakthroughs, innovations, and discussions shaping the world of AI.

June 30, 2023

GitHub Copilot: The AI Tool That Could Boost Global GDP by $1.5 Trillion

A recent study by GitHub found that GitHub Copilot could boost global GDP by $1.5 trillion by 2030. The study found that GitHub Copilot could help developers write code more quickly and efficiently. This would free up developers’ time to focus on other tasks, such as designing and testing code.

The study also found that GitHub Copilot could help to close the skills gap in the tech industry. By making it easier for developers to write code, GitHub Copilot could make it more accessible to people who do not have a traditional computer science background.

From techxplore

Microsoft Launches First Professional Certificate in Generative AI

Microsoft has announced the launch of its first professional certificate in generative AI. The certificate, which is available on LinkedIn Learning, provides learners with the skills they need to build and deploy generative AI models.

From zdnet

AI-Generated Drug Enters Clinical Trials, Ushering in a New Era of Drug Discovery

Insilico Medicine, a biotech company that uses artificial intelligence (AI) to discover and design new drugs, has announced that its first AI-generated drug has entered clinical trials in human patients. The drug, INS018_055, is an anti-fibrotic agent that is being developed to treat idiopathic pulmonary fibrosis (IPF), a progressive lung disease that has no known cure.

INS018_055 was discovered and designed using Insilico’s AI platform, which is trained on a massive dataset of drug molecules and biological data. The platform can identify potential drug targets and generate new drug candidates that are more likely to be effective and safe than those discovered using traditional methods.

The clinical trial of INS018_055 is a Phase 2 study that will enroll up to 120 patients with IPF. The study will assess the safety and efficacy of the drug in improving lung function and quality of life.

The success of INS018_055 could pave the way for the development of other AI-generated drugs. AI has the potential to revolutionize drug discovery, making it faster, more efficient, and more cost-effective. This could lead to the development of new treatments for a wide range of diseases.

From cnbc

ChatGPT on iOS Gets a Major Update with Bing Integration

OpenAI’s ChatGPT app for iOS has been updated with a major new feature: Bing integration. This means that paid subscribers can now use Bing search to give even more accurate answers to their questions. For example, if you ask ChatGPT “What is the latest news about the war in Ukraine?”, it will now be able to surface the most up-to-date information from Bing.

The integration with Bing is a significant upgrade for ChatGPT, as it allows the app to access information from a wider range of sources. This is especially important for users who are interested in staying up-to-date on current events.

To enable Bing integration in the ChatGPT app, simply open the settings and enable the “Browsing” option. Then, select the “GPT-4” model and choose “Browse with Bing” from the dropdown menu.

Bing integration is currently only available for paid subscribers of the ChatGPT Plus plan. However, OpenAI has said that it plans to make the feature available to free users in the future.

From theverge

June 29, 2023

OpenAI Sued for Scraping Private Data to Train ChatGPT

OpenAI, the company that created the popular chatbot ChatGPT, is being sued for scraping vast amounts of private data from the internet to train its AI model. The lawsuit alleges that OpenAI violated the privacy of millions of people by collecting their personal information without their consent.

The lawsuit was filed by a California law firm on behalf of a class of individuals who claim that their data was used by OpenAI without their knowledge or permission. The lawsuit alleges that OpenAI scraped data from a variety of sources, including social media posts, blog articles, and Wikipedia entries. The lawsuit also alleges that OpenAI used this data to train ChatGPT, which is a chatbot that can generate text that is indistinguishable from human-written text.

The lawsuit seeks damages for the alleged violations of privacy and seeks to prevent OpenAI from continuing to scrape private data.

From fortune

Microsoft’s AI-Powered Shopping Tools Make It Easier to Find the Best Deals

Microsoft has announced new AI-powered shopping tools for Bing and Edge that make it easier to find the best deals. The new tools include:

• Price Comparison: This tool allows you to compare prices from different retailers for the same product.
• Deal Finder: This tool helps you find current deals on products you’re interested in.
• Coupon Finder: This tool helps you find coupons that can be applied to your purchases.
• Product Recommendations: This tool provides personalized recommendations for products based on your past purchase history and browsing activity.

The new AI-powered shopping tools are available now in the Bing and Edge browsers.

From techcrunch

June 28, 2023

OpenAI Opens London Office, Heating Up AI Race in Europe

OpenAI has opened its first international office in London. The office will focus on research and engineering, and will help OpenAI to attract top talent from around the world.

The opening of the London office is a significant move for OpenAI, as it brings the company into direct competition with DeepMind, Google’s AI research lab. DeepMind is already well-established in London, and has been a major player in the development of AGI.

The opening of the London office is a sign that OpenAI is serious about its mission to ensure that AGI benefits all of humanity. By expanding to Europe, OpenAI is hoping to tap into the region’s rich talent pool and to contribute to the development of AGI in a global context.

From techcrunch

GPT-3 Can Spread Disinformation Better Than Humans

A new study has found that GPT-3, a large language model, can produce more compelling disinformation than humans. The study, which was conducted by researchers at the University of Milan, found that humans were unable to distinguish tweets generated by GPT-3 from tweets written by human users.

The study’s findings raise concerns about the potential for GPT-3 to be used to spread disinformation. The model’s ability to produce accurate information that is easy to understand makes it a powerful tool for spreading false information.

From techxplore

AI-Generated Nonsense Takes Over Amazon Bestseller Lists: What’s the Deal?

A recent report by Vice has found that a bunch of AI-generated books of nonsense are currently topping Amazon’s bestseller lists. The books, which are mostly self-published, are full of gibberish and nonsensical text, and they are often hard to read. It’s unclear how these books have managed to reach the top of the bestseller lists, but it’s likely that they’re being bought up by bots or other automated systems.

From vice

MosaicML Releases MPT-30B: The First LLM Trained on NVIDIA H100 GPUs

MosaicML has just released their MPT-30B, a large language model (LLM) that is trained on 1T tokens of text and code. MPT-30B is the first LLM to be trained on NVIDIA H100 GPUs, and it is available under the Apache 2.0 license. This means that it can be used for commercial purposes, and it is also compatible with the Hugging Face ecosystem.

From marktechpost

June 27, 2023

NVIDIA and Intel Lead the Way in GPT-3 Benchmarks, While Others Lag Behind

From forbes

ChatGPT Can Now Search the Web, But Only With Bing

OpenAI’s ChatGPT mobile app can now search the web, but only via Bing. This new feature, called Browsing, is available to subscribers of ChatGPT Plus. Browsing allows ChatGPT to access information that is beyond its original training data, such as current events and other information that is not yet publicly available. This makes ChatGPT a more useful assistant, particularly for research.

From techcrunch

Google’s Gemini AI to Surpass ChatGPT in Power and Versatility

Google has announced that its next AI project, Gemini, is set to be more powerful than ChatGPT. Gemini is a multimodal AI, meaning that it can understand and respond to a variety of data types, including text, images, and audio. This makes it more versatile than ChatGPT, which is only able to understand and respond to text.

Google claims that Gemini will be able to generate more creative and informative text, and that it will be better at understanding and responding to complex questions. The company is still in the early stages of developing Gemini, but it is expected to be released in the near future.

From cointelegraph

Alphabet’s AI Ambitions Are Too Ambitious, Bernstein Warns

Bernstein downgraded Alphabet on Tuesday, citing concerns about the company’s rapid pace of AI development. The investment firm warned that Alphabet is moving too fast in AI, and that this could lead to problems down the road.

Bernstein pointed to Alphabet’s recent investments in AI as evidence of its aggressive approach. The company has invested heavily in AI research, and has also acquired several AI startups. Bernstein is concerned that Alphabet is not taking the time to properly test and deploy its AI systems.

From yahoo

Baidu’s Ernie 3.5 Beats ChatGPT on Some Metrics, Claims to Be More Efficient

Baidu, the Chinese tech giant, has announced that its new AI model, Ernie 3.5, has outperformed OpenAI’s ChatGPT on some metrics. Ernie 3.5 is trained on a massive dataset of text and code, and is able to generate text, translate languages, and answer questions in a comprehensive and informative way.

According to Baidu, Ernie 3.5 has surpassed ChatGPT in comprehensive ability scores, and has outperformed GPT-4 in several Chinese capabilities. Baidu also claims that Ernie 3.5 is more efficient to train and run, which could make it more scalable for commercial applications.

From reuters

Monocle: The Tiny, Open-Source AR Lens That Puts AI in Front of Your Eye

Brilliant Labs has developed a tiny, open-source AR lens that puts AI in front of your eye. The device, called Monocle, can be clipped to eyeglasses or held to your face. It features a camera, microphone, and display, as well as a field-programmable gate-array processor (FPGA) that can be used to run AI models.

Monocle is currently in development, but the company has released a development kit that allows developers to start building AR applications. The kit includes the Monocle hardware, as well as a drag-and-drop interface called StreamLogic that makes it easy to create AR experiences.

From techcrunch

June 26, 2023

AI interviews will be used by 43% of companies by next year

A new survey found that 43% of companies plan to use AI interviews by next year. This is a significant increase from the current use of AI interviews, which is only 10%.

The survey, conducted by ResumeBuilder.com, also found that two-thirds of companies believe that AI interviews will improve hiring efficiency. Additionally, 15% of companies said that they plan to use AI interviews to make hiring decisions without any human input.

From foxbusiness

Nvidia and Snowflake Unite to Harness the Power of AI

Nvidia and Snowflake have announced a partnership that will bring Nvidia’s AI computing platform to Snowflake’s cloud data platform. This partnership will allow businesses to use their own data to train and deploy AI models on Snowflake’s platform.

The partnership is a win-win for both companies. Nvidia gains access to Snowflake’s vast customer base, while Snowflake gains access to Nvidia’s powerful AI computing platform. This partnership is a major step forward in the adoption of AI in the cloud.

From yahoo

DeepMind’s Sparrow Chatbot Set to Rival ChatGPT

DeepMind, the AI division of Google, is developing a new chatbot called Sparrow that is said to rival OpenAI’s ChatGPT. Sparrow is said to be able to cite sources through reinforcement learning, which would make it more reliable and informative than ChatGPT. Additionally, Sparrow is designed to be more helpful, correct, and harmless than other language models.

Demis Hassabis, the CEO of DeepMind, has warned that AI is “on the cusp” of reaching a level that could cause significant damage to humanity. However, he believes that Sparrow could be used to help build safer and more useful artificial general intelligence (AGI).

From techcrunch

Databricks acquires MosaicML to bolster its AI and ML capabilities

Databricks, a leading provider of data and AI cloud platform, has acquired MosaicML, a company that develops open-source tools for machine learning. The acquisition is worth $1.3 billion.

MosaicML’s tools are used by data scientists and engineers to build and deploy machine learning models. The company’s products include Mosaic, a framework for building and managing machine learning pipelines, and MLflow, a platform for tracking and managing machine learning experiments.

Databricks says that the acquisition of MosaicML will help it to accelerate its growth in the AI and ML space. The company plans to integrate MosaicML’s tools into its own platform, making it easier for users to build and deploy machine learning models.

From techcrunch

AI and misinformation: A recipe for disaster in the 2024 elections

Former Google CEO Eric Schmidt has warned that the 2024 elections are going to be a “mess” due to the use of AI to spread misinformation. Schmidt said that AI-generated deepfakes and other forms of disinformation could be used to sway voters and undermine the electoral process.

Schmidt’s warning is timely, as there is already evidence that AI is being used to spread misinformation in the political sphere. For example, in the 2020 US presidential election, there were reports of AI-generated deepfakes being used to spread false information about both candidates.

From cnbc

AI can detect cavities and gum disease more accurately than humans

Artificial intelligence (AI) is being used to help dentists catch more cavities and gum disease. AI is more accurate than humans at detecting these conditions, and it is also unbiased, which means that it is not influenced by the dentist’s personal biases. This can lead to earlier diagnoses and more effective treatments.

In addition to detecting cavities and gum disease, AI is also being used to help dentists with other tasks, such as planning treatment and predicting treatment outcomes. As AI continues to develop, it is likely to play an even greater role in dentistry in the future.

From foxnews

June 25, 2023

Midjourney Released Model Version 5.2

Midjourney, a text-to-image AI, has released its latest model version, 5.2. This new version includes a number of improvements, including:

• Improved image quality: The new model produces more detailed, sharper images with better colors, contrast, and compositions.
• New Zoom Out feature: This feature allows users to zoom out of an image, generating new versions with different camera angles.
• Shorten command: This command helps users to shorten their text prompts, making it easier to generate images with specific keywords.

From midjourney

Every week, we’ll meticulously curate a selection of stories from top AI media outlets and distill them into a digestible format, ensuring you stay up-to-date with the latest developments without having to spend hours browsing the web. From groundbreaking research to real-world applications, ethical debates to policy implications, AI Weekly Digest will be your essential guide to navigating the ever-evolving landscape of artificial intelligence. Join us on this exciting journey as we explore the future of AI together, one week at a time.