Samwise Tech/AI/Robotics Newsletter
Thursday, July 16, 2026
Thinking Machines Releases Inkling, a 975B-Parameter Open-Weight Multimodal Model
Thinking Machines Lab, the startup founded by former OpenAI CTO Mira Murati, released its first in-house AI model on July 15: Inkling, an open-weight multimodal system. The mixture-of-experts architecture carries 975 billion total parameters but activates only about 41 billion for any given task. Trained on 45 trillion tokens spanning text, image, audio, and video, Inkling reasons natively across all four modalities. The company designed Inkling to give calibrated answers, flag uncertainty, and let users dial thinking effort up or down for speed. Thinking Machines explicitly notes Inkling is “not the strongest overall model available today” and built it to answer directly on topics that may be subject to censorship.
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
China Approves Apple Intelligence Powered by Alibaba’s Qwen AI Across All Apple Platforms
China’s Cyberspace Administration of China approved Apple’s AI services on July 15, enabling Apple Intelligence to launch in the country through a partnership with Alibaba. The deal integrates Alibaba’s Qwen AI model into Apple’s operating systems — iOS, iPadOS, macOS, and visionOS. Alibaba confirmed that Qwen would be “integrated into Apple Intelligence experiences” but provided no launch timeline. China is a critical market for Apple; the company reported a 28 percent increase in Greater China sales to $20.5 billion in its most recent quarter. The approval marks a significant expansion of generative AI across Apple’s global device base and ends a lengthy wait for both companies.
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
Anthropic and Blackstone Launch $1.5B Enterprise AI Implementation Firm Ode
Anthropic and private equity firms including Blackstone, Hellman and Friedman, and Goldman Sachs launched Ode with Anthropic, a $1.5 billion AI implementation company focused on deploying engineers directly inside enterprise clients. Currently employing 100 engineers, Ode operates on a “Claude-first” principle but will use rival models when the task demands it. The company originated from Blackstone’s experience identifying a deployment gap while rolling out AI across its portfolio, and subsequently acquired AI engineering services startup Fractional AI. Additional investors include Apollo Global Management, General Atlantic, GIC, Leonard Green, and Sequoia Capital. Ode’s bet is that helping businesses apply AI models is the next trillion-dollar category.
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
Microsoft Uses AI to Find a Record 570 Security Flaws in July Patch Tuesday
Microsoft’s July 2026 Patch Tuesday addressed a record 570 security vulnerabilities, with the company crediting AI for discovering many previously unknown bugs. Among the patched flaws were at least two zero-days exploited before disclosure: one allows privilege escalation to system administrator in Windows Server, and another affects SharePoint, which CISA flagged as under active exploitation. A flaw in the remastered Age of Empires II allowed attackers to take over systems via a malicious game invite. Windows chief Pavan Davuluri noted that “as AI helps defenders discover more issues, customers will see a higher volume of security updates included in each security release.”
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
Toyota-Backed Walden Robotics Emerges from Stealth at $1.1B Valuation
Walden Robotics, a Cambridge, Massachusetts startup spun out of Toyota Research Institute in January, emerged from stealth on July 15 with approximately $300 million in funding at a $1.1 billion valuation. The seed round was co-led by Deviation Capital and Toyota Motor Corp., with participation from Nvidia, Boeing, AE Ventures, Samsung Ventures, Prologis Ventures, and CoreWeave Ventures. Walden’s robots feature a humanoid torso with two arms and a wheeled base, chosen to ease factory safety certification. Since February, the robots have operated in production manufacturing and logistics tasks at a Toyota plant in North America, moving from first pilot to full deployment in under two months.
Sources: The Robot Report Share ↗ ✉︎ Email 💬 Text
OpenAI Launches $230 Codex Micro Keyboard for AI Coding Agents Amid Apple Lawsuit
OpenAI entered the physical hardware market on July 15 with the Codex Micro, a $230 light-up keyboard co-designed with specialty maker Work Louder to pair with its Codex AI coding agent. The Micro features Agent Keys that display real-time agent status, customizable Command Keys for frequent Codex actions, a joystick for launching workflows, and a dial that controls how much reasoning an agent applies to a task. The launch arrives amid a lawsuit from Apple, filed July 10, alleging that OpenAI deliberately extracted confidential Apple hardware information; OpenAI has denied wrongdoing. OpenAI described the Micro as a limited-run collaboration, signaling it is more proof of concept than mass-market product.
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
India’s Emergent Hits $1.5B Unicorn Valuation with $130M Series C for AI Coding Platform
Indian AI coding startup Emergent raised $130 million in a Series C round on July 15, reaching a $1.5 billion post-money valuation — a five-fold increase in six months. The round was led by private equity firm Creaegis, with new backers MNI Ventures-Claypond and Sentinel Global joining existing investors Khosla Ventures, SoftBank’s Vision Fund 2, Lightspeed, and Y Combinator. The deal brings Emergent’s total funding to $230 million; the company had raised a $70 million Series B at a $300 million valuation in January. Emergent plans to use the capital to improve app success rates and expand support for complex AI agent workflows using local and open-source models.
Sources: TechCrunch Share ↗ ✉︎ Email 💬 Text
Tech Pulse
Top Frontier Models (SWE-bench V): GPT-5.5 (88.7%) | Claude Opus 4.8 (88.6%) | Gemini 3.1 Pro (80.6%)
Top Open Source Models (MMLU): Llama 4 Maverick (85.5%) | Qwen 3 235B (85.2%) | DeepSeek V3.2 (84.8%)
Top Small Models 15–50B (MMLU): Phi-4 14B (84.8%) | Qwen3.5-32B (82.1%) | Mistral Medium 3.5 (81.0%)
Top Edge Models 0–15B (MMLU): Qwen 3.5 9B (79.4%) | Granite 4.1 8B (78.9%) | Gemma 3 4B (73.6%)
AI Leaders (market cap): NVIDIA $5.1T | Alphabet $4.3T | Apple $4.3T
Robotics Leaders: Intuitive Surgical $175B | ABB $165B | Figure AI $39B (private est.)
Curated by JD · samwise.agency