All in: South Korean government to invest $22 billion in AI and chips

In today’s newsletter:

• ABA Podcast EP2: Interview with Google Gemini and Vertex AI Product Manager

• Insights: Top three LLM frameworks to help you develop GenAI apps

• What can AI startup investors learn from the Black Myth: Wukong game development journey?

• Tutorial: The Power of Custom Instructions

• 4 AI tools in the spotlight

Read time: 4-5 mins

🎙️ ABA Podcast #EP2: Google Gemini Group Product Manager on “how close are we to building an enterprise-ready AI experience?”

In this episode, Leo interviews Lewis Liu, Group Product Manager at Google, Gemini and Vertex AI, about the impact of generative AI models on user experience and enterprise AI readiness.

The discussion centers on the transformative impact of generative AI models on enterprise readiness and user experience. Key topics include the rapid evolution of generative AI, the mix of experts architecture, and the role of language models (LLMs) in enterprise applications.

Some key questions addressed:

1. How to ground AI so that it works for us, instead of working for itself

2. What does the enterprise-ready AI experience look like?

3. Should we be chasing quality benchmarks/leaderboards?

4. What actually matters for optimising model capabilities?

Chapters:

• 04:28 The Journey of Generative AI in Google

• 07:13 Automating Code Generation and Completion

• 09:48 Unlocking Limitations with Sensitive Data

• 12:53 The Promise of Mix of Experts Architecture

• 18:22 The Role of Data in the Progression of LLMs

• 20:13 Adding Human Expertise to Improve Factuality

• 23:50 Advantages of First-Party Data and Infrastructure Investment

• 26:38 Grounding AI in Factual Information

• 30:00 Building an Enterprise-Ready AI Experience

• 32:07 Improving Factuality and Contextuality in Model Responses

• 35:09 The Role of Google Search in Language Models

• 36:57 Handling Conflicting Data Signals in LLMs

• 40:17 LLMs in Startups and Enterprise Use Cases

• 49:48 The Commoditization of LLMs and Differentiated Segments

• 55:54 Changing Ways of Knowledge Acquisition

• 01:03:13 The Identity of LLMs and the Need for National Boundaries

Make sure you’re subscribed to get notifications → Youtube | Spotify | Apple  | RSS 

The current landscape of enterprise adoption of Artificial Intelligence (AI) reflects a significant shift towards more definitive integration into business operations. Recent surveys indicate that 77% of companies are either using or exploring AI, with 83% considering it a top priority in their strategic plans.

In this article, we deep dive into the three most popular LLM frameworks that have generated notable traction in supporting businesses developing GenAI Apps with a case study. 

Your choice of LLM frameworks should always depend on your specific need. To find out more about the differences between key features and capabilities between the three frameworks, and correctly choose the fit-for-purpose framework for your GenAI development:

Check out this article → here.

We’ve partnered with an expert guest author, Stig Korsholm, to bring you more in-depth tutorials on prompt engineering. Today’s tutorial is a follow-up from last week’s “Comparative guide to 9 Prompt Engineering Frameworks for Tech Professionals”.

This article aims to unravel how custom instructions can enhance the capabilities of AI, offering a more tailored and efficient interaction, and thus serve as a vital complement to the principles of prompt engineering previously discussed.

Check out this tutorial → here.

News from Asia:

• South Korea government plans to allocate $22.6 billion to invest in AI and chips over the next five years.

• AWS to invest $6.2 billion to build data centers in Malaysia, tripling Microsoft and Google Cloud’s respective Malaysian investments.

• Black Myth: Wukong game sold 10 million copies in three days - what does it tell us about in investing AI startups?

• Chinese and US scientists create AI model to help develop new drugs.

• Hitachi and Singtel to deepen AI data center collaboration, combining Hitachi’s power and cooling equipment with Singtel’s cloud technology.

• Chinese entities turn to Amazon cloud and its rivals to access high-end US chips.

News from the West:

• OpenAI, Adobe and Microsoft have thrown their support behind a California bill AB 3211 requiring tech companies to label AI-generated content.

• Apple expected to debut first generative AI iPhone (16) at its September 9 event.

• Three of the five cofounders of French AI startup H leaves pre-product and just three months after $220 million seed round. (We wish we were joking)

• Viggle AI, a genAI character animation startup, secures $19 million in Series A funding led by Andreessen Horowitz

• Reliant AI, a genAI-powered data analytics software provider, announced its launch out of stealth with $11.3 million in seed funding, co-led by Tola Capital and Inovia Capital.

Trending Tools & Apps

1. D-ID is the latest company to ship a tool for translating videos into other languages using AI technologies. However, in this case, D-ID also clones the speaker’s voice and changes their lip movements to match the translated words as part of the AI editing process.

2. ShellMate is a lightweight, open-source app that gives you superpowers in your Mac Terminal. It will observe what you're doing and detect errors, suggest fixes, or even suggest the next command you haven't run yet but should.

3. xGen-VideoSyn-1, Salesforce’s text-to-video (T2V) model that generates realistic scenes from textual descriptions

4. Claude’s system prompts are published by Anthropic for all to see. Transparency FTW!