What a week! IBM watsonx and Google AI announcements

Hey Readers, Happy Sunday! 👋

In today's edition, I'll cover the following:

1- My concept of the week: The AI Industry Catalog

2- AI News of the Week

3- My new Interview: A demo of the Predictive Maintenance Use Case with Yana, a Senior Data Scientist from IBM

Let's dive in! 🤿

📚 My AI posts this Week: AI For Drug Discovery

In recent years, the field of AI has made significant strides in numerous domains, one of which is healthcare. AI has the potential to revolutionize the drug discovery and development process, leading to more effective and cost-efficient treatments for patients. In this blog post, we will deep dive into the role AI plays in the drug discovery and development process and explore some of the major breakthroughs and challenges faced in this field.

AI played a big role in the drug discovery process for Covid-19 and IBM recently announced a partnership with Moderna to collaborate in drug discovery using Quantum and AI technology. Don’t miss this new exciting use case!

🗞️Top News of the Week

1. IBM announced the next generation of AI Platform

Despite the growth in AI adoption, scaling and operationalizing AI in business remains a challenge, with only 54% of AI projects progressing from pilot to production. IBM aims to facilitate AI utilization by creating products that adapt to organizations' data, thereby providing personalized AI models and a seamless experience. At Think 2023, IBM announced the release of watsonx, an AI and data platform that combines traditional machine learning with generative AI capabilities. This will allow businesses to train, adjust, and implement AI models on a large scale, transparently and responsibly.

Watsonx integrates into IBM software products across various sectors, including digital labor, IT automation, security, sustainability, and application modernization, unlocking new levels of productivity and value for clients. Watsonx comprises three products that aim to simplify the AI experience for businesses:

1. Watsonx.ai: An enterprise studio for AI creators to train, test, adjust, and implement traditional machine learning and new generative AI capabilities. It is expected to be generally available in July.

2. Watsonx.data: A data store that provides businesses with a single access point to all their data across all clouds and on-premise environments. It is expected to be generally available in July.

3. Watsonx.governance: An end-to-end toolkit for data and AI governance that enables transparent and explainable AI workflows. It is expected to be generally available in October.

IBM states that the watsonx platform will transform businesses by scaling AI and unlocking new value with foundational models and generative capabilities. IBM also announced a big partnership with HuggingFace.

2.  Google is going all in with AI

Google recently unveiled several new AI developments at Google I/O:

1. New language models PaLM 2 and Gemini: PaLM 2, which reportedly outperforms GPT-4 in various tasks, is now available. It comes in different sizes named after animal constellations: Gecko, Otter, Bison, and Unicorn. Gemini, a new multimodal LLM, is under development.

2. Bard updates: Google's answer to ChatGPT, Bard, is now available for public use and runs on PaLM 2. It can be exported to various applications like Google Docs, Sheets, Replit, and Gmail. New "tools" in Bard will enable users to access features like Google Maps and Adobe Firefly.

3. Enhanced search capabilities: Google Search will now feature AI-generated snapshots that answer queries using content from top links, accompanied by traditional links and sponsored ads. Users can ask follow-up questions in conversation mode.

4. Workspace AI sidekick (Duet AI): This feature offers automatic prompt suggestions and can perform various tasks like converting text to tables in Sheets, finding information in Gmail threads, and creating images in Slides. It also integrates with Gmail and Docs to provide writing assistance. This is in response to Microsoft Copilot.

5. Additional updates: Other announcements include a Magic Editor in Google Photos, public text-to-music tools, upcoming AI capabilities on Android, AI improvements for Chrome, Maps, and ML Hub for developers to train and deploy models.

3. Anthropic announced 100k Context Windows

Anthropic has developed a new technique called "100k context windows" for training AI models on large datasets of text and code. This comes with an update of its AI chatbot, Claude, to process up to 75,000 words of context, which is three times more than OpenAI's chatGPT and GPT-4 capacity. This enhancement enables Claude to quickly digest, summarize, and reference information from hundreds of pages.

To demonstrate its capabilities, Claude was fed an 85-page Netflix corporate filing, and it successfully highlighted important terms for investors, created tables from Netflix's balance sheet, and provided analyses on Netflix's market positioning. These tasks, which would typically take a human analyst several days, were accomplished by Claude in seconds.

This upgrade has broad implications, potentially eliminating the need for hours spent sifting through annual reports, legal documents, codebases, and legislation. In the future, Claude might even generate graphs, images, and PowerPoint presentations from a PDF, potentially increasing an analyst's productivity tenfold.

This update puts Anthropic in direct competition with OpenAI and threatens various "chat with your PDF" tools. It remains to be seen if OpenAI will respond by increasing its own AI's context window. You can see a demo below:

4. AI for Advertisement with Meta’s AI Sandbox

Meta is introducing new generative AI features for advertisers. These features can help advertisers to create more personalized and engaging ads, save time and money, and improve the quality of their ads.

These features, which are still in beta, allow advertisers to create more personalized and engaging ads.

• One of the new features is called Creative Templates. Creative Templates are pre-made ad formats that can be customized with text, images, and videos. This makes it easy for advertisers to create high-quality ads without having to start from scratch.

• Another new feature is called Creative Suggestions. Creative Suggestions are personalized recommendations for ad creatives based on the advertiser's target audience and campaign goals. This can help advertisers to create ads that are more likely to resonate with their target audience.

• Meta is also introducing a new tool called Creative Hub. Creative Hub is a central place for advertisers to manage their creative assets and create new ads. It includes features such as a library of pre-made assets, a design tool, and a collaboration platform.

5. Web Browsing and Plugings for ChatGPT Plus

OpenAI is rolling out web browsing and Plugins to all ChatGPT Plus users over the next week! Moving from alpha to beta, they allow ChatGPT to access the internet and use 70+ third-party plugins. Plugins are like a marketplace for ChatGPT users to extend functionality.

See you next Sunday!

Armand 😎