Mira Murati’s Thinking Machines Lab 💡, Grok 3 ⚡, IBM's $7B Bet on Vertical AI 💰

💬 OpenAI Developments

Thinking Machines Lab is ex-OpenAI CTO Mira Murati’s new startup

Mira Murati, former Chief Technology Officer of OpenAI, has launched a new AI startup called Thinking Machines Lab, aiming to make AI systems more accessible and customizable. The company has recruited a team of approximately 30 researchers and engineers, including former OpenAI co-founder John Schulman as chief scientist and ex-OpenAI researcher Barret Zoph as Chief Technology Officer.

Introducing the SWE-Lancer benchmark

OpenAI has introduced SWE-Lancer, a benchmark comprising over 1,400 freelance software engineering tasks sourced from Upwork, collectively valued at $1 million USD. This benchmark includes tasks ranging from $50 bug fixes to $32,000 feature implementations, as well as managerial tasks where models select between technical proposals. Evaluations reveal that current frontier models are unable to solve the majority of these tasks, highlighting the challenges in applying AI to real-world software engineering. To support further research, OpenAI has open-sourced a unified Docker image and a public evaluation split named SWE-Lancer Diamond.

🌐 DeepSeek Developments

Baidu adopts DeepSeek AI models, chasing Tencent in race to embrace the hot start-up

Baidu is integrating DeepSeek's AI models into its search engine, following Tencent's recent adoption of DeepSeek's technology in Weixin's search feature. This move highlights the rapid embrace of DeepSeek's innovations by major Chinese tech firms, aiming to enhance their AI capabilities and user experiences.

AI start-up DeepSeek expands business scope in potential shift towards monetisation

DeepSeek, a Chinese AI start-up known for its open-source large language models, has expanded its business scope to include "internet information services," indicating a potential shift towards monetisation. This move suggests the company's intention to transition from a research-focused approach to developing a sustainable business model.

🚀 Tech Industry Moves

Elon Musk's AI startup, xAI, has unveiled Grok-3, a new chatbot integrated with X (formerly Twitter), aiming to compete with AI models from DeepSeek, OpenAI, and Google. Grok-3, available to Premium+ subscribers, boasts over ten times the computing power of its predecessor and introduces advanced reasoning capabilities, including a "Big Brain" mode for complex tasks. Additionally, xAI has launched DeepSearch, a reasoning-based chatbot designed to enhance research and data analysis.

Meta has announced two major events for 2025: LlamaCon on April 29, a developer conference focused on the latest in Meta’s open-source Llama AI models, and Meta Connect on September 17-18, showcasing innovations in Meta Horizon, XR, and AI glasses. More details on both events will be shared in the coming weeks.

Microsoft has announced a $1 million investment to train one million Nigerians in artificial intelligence (AI) skills over the next two years. This initiative aims to equip Nigerians with AI expertise, fostering digital transformation and enhancing workforce readiness in the country.

Microsoft has unveiled Muse, a generative AI model developed in collaboration with Ninja Theory, designed to generate gameplay sequences based on player interactions and visuals. Trained on extensive human gameplay data from "Bleeding Edge," Muse aims to assist game developers in prototyping, preserving classic games, and enhancing game design processes.

Nvidia and the Arc Institute have introduced Evo 2, an AI model trained on nine trillion DNA base pairs from over 128,000 species, designed to predict functions, analyze mutations, and design new genetic sequences. This advancement builds upon previous efforts, such as the original Evo model developed by the Arc Institute, Stanford University, and UC Berkeley, which analyzed 2.7 million microbial and viral genomes to create novel gene-editing proteins.

Nvidia: A recent Nvidia report reveals that 97% of telecom providers are adopting AI technologies, with 84% reporting increased annual revenue due to AI-driven efficiencies. The study highlights AI’s role in optimizing network performance, improving customer service, and enhancing operational automation, positioning AI as a crucial factor in the telecom industry's future growth.

Nvidia's stock has shown resilience, nearly recovering from a significant 17% drop on January 27, 2025, caused by concerns over Chinese firm DeepSeek's cost-effective AI model. The rebound is bolstered by substantial AI infrastructure investments from U.S. tech giants and international commitments, such as South Korea's plan to acquire 10,000 Nvidia GPUs for a national AI computing center.

💡 Emerging AI Technologies and Companies

Ant Group enters China's growing humanoid robot industry amid rising tech interest

Ant Group, the fintech giant affiliated with Alibaba, has entered China's burgeoning humanoid robot industry by establishing its subsidiary, Shanghai Ant Lingbo Technology, in late 2024. This move aligns with China's strategic focus on humanoid robots to drive economic growth and technological advancement.

IBM’s $7B Bet on Vertical AI and What It Means for SaaS Founders

IBM is investing $7 billion annually in Vertical AI, focusing on industry-specific AI solutions through its WatsonX platform. This shift presents a major opportunity for SaaS founders to build specialized AI applications leveraging enterprise data, which accounts for less than 1% of public foundation model data but delivers higher value. IBM’s Instruct Lab approach reduces AI costs by 98.5% and saves 35% in time, challenging the trend of larger models. Strategic partnerships with ServiceNow, Adobe, and Salesforce are accelerating enterprise AI adoption, signaling a shift towards more efficient, domain-focused AI development.

🎨 AI in Creative Industries

Adobe announced new AI-powered features for Lightroom and Camera Raw at Max Japan 2025. Adaptive Profiles dynamically adjust tones and colors for a customized HDR editing experience, while Distraction Removal (currently in Early Access) detects and removes unwanted elements like reflections and photobombers. Find Duplicates, available in Lightroom Web, helps clean up libraries by removing duplicate and blurred images. Additionally, Lightroom mobile for iOS now offers faster auto-generated shareable links for easier photo sharing, with updates for Android users expected later.

If you want more updates related to AI, subscribe to our Newsletter