Generative AI News Rundown - LLM Palooza with Llama, Mistral, Phi & Grok, Plus New Funding, Adobe, Apple, and More - Voicebot Podcast Ep 379

2024-06-17 · Source: The Voicebot Podcast · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Cloud Computing & IT Infrastructure · Depth: Intermediate, extended

Summary

The generative AI landscape experienced significant activity over two weeks, highlighted by Meta's Llama 3 launch, featuring 8B and 70B parameter models with a 400B model anticipated, demonstrating substantial performance gains over Llama 2. Other notable model releases included X.ai's Grok-1.5V multimodal model, Mistral's efficient 8x22B LLM, and Microsoft's Phi AI models, which are small enough to run on phones. Enterprise adoption data from Andreessen Horowitz indicates 93% of companies use multiple models, with OpenAI dominating production use at nearly 70% market share. Control, customizability, and cost are key drivers for open-source adoption, with 60% of enterprises prioritizing control. The period also saw nine-figure funding rounds for Cognition Labs ($175M at $2B valuation) and Augment ($252M), a $700M acquisition of Run.ai by Nvidia, and Perplexity's enterprise debut alongside a $1B valuation. Apple acquired French AI startup Datakalab, reinforcing its on-device AI focus.

Key takeaway

For CTOs and VP of Engineering evaluating generative AI solutions, the rapid proliferation of high-quality, efficient models like Llama 3 and Microsoft Phi necessitates a flexible, multi-model strategy. Your teams should prioritize open-source options for greater control and cost-efficiency, especially as smaller models achieve "good enough" performance for many use cases. This shift reduces reliance on single vendors and mitigates supply chain risks for GPUs, enabling broader deployment across diverse platforms, including edge devices.

Key insights

The generative AI market is rapidly diversifying with new models, significant investments, and a strong enterprise shift towards multi-model and open-source strategies.

Principles

• Enterprise AI adoption favors multi-model strategies.
• Control and customizability drive open-source preference.
• Data quality and quantity are critical for model performance.

Method

Microsoft's Phi models demonstrate that high-quality, curated training data (e.g., "textbooks are all you need") can achieve GPT-3.5 level performance with significantly fewer parameters, enabling efficient small language models.

In practice

• Evaluate multiple LLMs for diverse use cases.
• Prioritize open-source models for control and customization.
• Consider small language models for on-device or cost-sensitive applications.

Topics

• Large Language Models
• Generative AI Adoption
• Llama 3
• Small Language Models
• AI Funding & Acquisitions

Best for: CTO, VP of Engineering/Data, Machine Learning Engineer, AI Engineer, AI Product Manager, Director of AI/ML

Related on AIssential

• LWiAI Podcast #234 - Opus 4.6, GPT-5.3-Codex, Seedance 2.0, GLM-5 — The AI sector is rapidly advancing with new models, generative media tools, and significant venture capital investments.
• Meta Llama 3.1 405B May Spark a Tectonic Shift in the LLM Market Adoption — Meta's Llama 3.1 405B is the first open model to achieve frontier-level performance, rivaling top proprietary LLMs.
• LLM Vs AI: A Practical Guide to Differences, Use Cases, and Tools — Generative AI and LLMs are transforming enterprise operations through diverse applications, requiring careful selection and deployment.
• Latest open artifacts (#17): NVIDIA, Arcee, Minimax, DeepSeek, Z.ai and others close an eventful year on a high note — The open model ecosystem is rapidly evolving with diverse architectures, data transparency, and specialized performance.
• Llama 3 on Your Local Computer | Free GPT-4 Alternative — Llama 3 achieves significant performance gains through extensive training data, an improved tokenizer, and advanced fine-tuning.
• Why Small Language Models (SLMs) Are Quietly Winning in 2026 — My Switch from Giant LLMs — Specialized small language models offer superior practicality, speed, and cost-efficiency over monolithic large models for most product use cases.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Voicebot Podcast.