Staff AI Engineer

Alphabet Guides AI Capex to $205 Billion as Free Cash Flow turns negative - July 23, 2026. Top Story. Alphabet Guides 2026 AI Capex to $205 Billion - and Free Cash Flow Just Went Negative Alphabet's Q2 2026 earnings, reported Wednesday, July 22, showed revenue up 24% year-over-year to $119.8 billion, with Google Cloud revenue exploding 82% to $24.8 billion. But the number that moved markets was capex guidance: Alphabet raised its expected 2026 AI infrastructure spend to as high as $205 billion, up from the $180-190 billion range it gave last quarter, and warned that 2027 spending will "increase significantly." Capex more than doubled year-over-year to $44.9 billion in the quarter alone, pushing free cash flow negative to the tune of $5.9 billion. CFO commentary reiterated that the company remains "in a supply-constrained environment." Why it matters: This is the clearest signal yet that the frontier-AI capex race is still accelerating, not plateauing - Alphabet is willing to run free cash flow negative and commit to even higher 2027 spending rather than slow down, a bet the market is still digesting (see chip-stock reaction below). Research & papers. Nothing independently verified as newly published in the last 24 hours met the bar for inclusion. A sweep of arXiv (cs.AI, cs.LG, cs.CL, cs.CV) and major lab research pages turned up nothing dated July 22-23 that wasn't either already covered in prior briefings or too preliminary to independently verify. Industry & startups. Anthropic Commits $200 Million to a New Economic Futures Research Fund, Opens Its Economic Index to Anyone on Claude Anthropic announced on July 22 a $200 million commitment to the Anthropic Economic Futures Research Fund, backing external research on interventions to prepare society for AI's economic disruption - prioritizing worker-level impacts, transition support, modernized income support, worker equity stakes, and evidence on public investment. Alongside it, Anthropic shipped an Anthropic Economic Index connector for Claude, letting anyone query the Index's AI-usage data directly inside a Claude conversation, no separate tool required. Why it matters: It converts Anthropic's economic-impact messaging into an actual funded research program and a public-facing data tool, rather than just talking points. China's PsiBot Hits $1.48 Billion Valuation in New Embodied-AI Funding Round PsiBot (also known as Lingchu Intelligence), a Chinese embodied-AI and robotics startup building vision-language-action models for other robot makers, is close to finalizing roughly $100 million in new funding at a $1.48 billion valuation, led by carmaker Chery Automobile and sensor maker Lens Technology. The company has now raised about $300 million since founding in 2024. Why it matters: Another concrete, large-dollar signal that Chinese capital is piling specifically into embodied/robotics AI, not just LLMs. UK's "Humanoid" Becomes Europe's First Pure-Play Humanoid Robotics Unicorn | Dated July 21 - outside the strict 24-hour window but not covered in any prior briefing. Humanoid, a UK-based robotics company, raised $152 million in Series A funding at a $1.35 billion post-money valuation, led by Prime Movers Lab with Schaeffler, Bosch, and others participating. The company builds wheeled humanoid robots and plans Beta deployments in Q4 2026 across logistics, manufacturing, and retail. Why it matters: A distinct data point from the Blackstone/Futronic actuator deal covered earlier this week - this is capital going into a full humanoid-robot company rather than a components supplier. Healthcare AI's Investor Narrative Has a Credibility Problem, New Survey Finds A Fishbone Advisors report published July 22 found 48-67% of institutional investors believe healthcare-company management overemphasizes AI relative to its actual financial contribution. Only 33% of healthcare-specialist investors (vs. 51% of generalists) expect AI to materially move healthcare valuations within three years. Why it matters: A concrete, numbers-based counterpoint to healthcare-AI hype from the investors actually pricing these companies. Tools & releases. OpenAI Launches Presence, an Enterprise Platform for Deployed Voice and Chat Agents OpenAI launched Presence on July 22 - an enterprise platform for deploying AI agents with defined knowledge access, system permissions, and human-escalation rules, plus a Codex-powered loop that reviews agent interactions and proposes improvements for staff to approve. Early customers include BBVA Mexico and SoftBank Corp. Why it matters: It's OpenAI's most concrete pitch yet for governed, production enterprise agents rather than open-ended assistants - with guardrails as the headline feature, not an afterthought. Global AI & geopolitics. Nothing genuinely new independently verified in the last 24 hours beyond what's already covered in recent briefings (Treasury Secretary Bessent's sanctions threat over alleged Chinese model distillation and China's reported consultations on model-weight export controls were both covered in yesterday's briefing). Energy, infrastructure & chips. Asian Chip Stocks Rally on Alphabet's Capex Optimism SK Hynix rose as much as 6.5% and MediaTek as much as 5.2% in Asian trading July 23 after Alphabet's raised AI capex guidance (see Top Story) reassured investors that hyperscaler AI spending isn't slowing. SK Hynix gets more than 7% of its sales from Google; MediaTek is reported to be developing custom processors for Alphabet. Why it matters: A direct, quantifiable read-through from one hyperscaler's spending guidance to the memory and chip-design supply chain underneath it. AI Agents & autonomy. Jack Dorsey's Block Launches Buzz, an Open-Source Workspace Where AI Agents Work Alongside Employees Block launched Buzz on July 21 - a free, open-source, Nostr-based group-chat platform combining channels, threads, DMs, voice, and code repos, where AI agents (running on any model or harness, including Claude Code, Codex, or Block's own goose) participate as workspace members with their own cryptographic identities, rather than as command-only bots. Built explicitly to reduce Block's reliance on Slack and GitHub. Why it matters: A concrete, shipped example of "AI agents as team members" architecture - decentralized and model-agnostic by design, rather than tied to one vendor's agent stack. Safety, alignment & ethics. Nothing independently verified as newly published in the last 24 hours from validated safety orgs or labs with a confirmed July 22-23 date. (OpenAI's long-horizon-model alignment writeup and Anthropic's agentic-misalignment case studies are both more than 24 hours old and were eligible for earlier briefings.) Numbers & signals. * $205 billion - Alphabet's raised 2026 AI capex guidance, up from $180-190B * $119.8 billion - Alphabet Q2 2026 revenue, up 24% YoY * 82% - Google Cloud revenue growth to $24.8 billion * -$5.9 billion - Alphabet's Q2 free cash flow (negative) * $200 million - Anthropic's new commitment to the Economic Futures Research Fund * $1.48 billion - PsiBot/Lingchu Intelligence's new valuation * $152 million / $1.35 billion - Humanoid's Series A raise / post-money valuation * +6.5% / +5.2% - SK Hynix / MediaTek share moves on Alphabet capex optimism * 48-67% - Share of institutional investors who see healthcare's AI narrative as overblown (Fishbone Advisors) Worth thinking about. Today's stories all point at the same tension from different angles: capital keeps flowing into AI infrastructure and embodied AI even as the people closest to the money get more skeptical about the payoff. Alphabet is willing to run cash flow negative and pre-commit to an even bigger 2027 budget. Chinese and European robotics startups keep raising at billion-dollar-plus valuations. And yet healthcare investors - arguably the most sophisticated buyers of "AI value" claims in any vertical - are telling surveyors the AI narrative in their sector is overstated relative to actual financial contribution. None of these are contradictions exactly; they're just evidence that the capex/hype cycle and the ROI-scrutiny cycle are now running in parallel, at the same time, in the same industry. Government & regulation. No major new US, EU, or other regulatory actions were independently verified as enacted in the last 24 hours. Frontier lab dispatch. Google - July 22: Reported Q2 2026 earnings; raised 2026 AI capex guidance to $205B and warned 2027 spend will increase significantly (see Top Story). Anthropic - July 22: Committed $200M to the Economic Futures Research Fund and launched the Economic Index connector for Claude. OpenAI - July 22: Launched Presence, an enterprise platform for governed voice/chat AI agents. Quick links. Tier 1 - Frontier Labs (Primary Sources) Tier 1 - Financial / Official Tier 3 - Tech & AI News Media

Humanoid raises $152 million and reaches unicorn status. With a valuation of 1.35 billion U.S. dollars, the British startup is one of Europe's most valuable robotics companies. Schaeffler, Bosch, and other investors are betting on humanoid robots for manufacturing and logistics. News by Marc Nemitz · London, 22. July 2026 The British robotics company Humanoid has closed a Series A funding round of $152 million, making it Europe's first pure-play humanoid robotics unicorn with a post-money valuation of $1.35 billion. Founded in 2024, the company has now raised a total of $270 million. The funding round was led by Prime Movers Lab. Other investors include Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures. Humanoid plans to use the capital to accelerate the development of its next generation of robots, prepare for industrial mass production, and further expand its AI platform. Humanoid robots for industry. Humanoid develops humanoid robots for industrial applications in manufacturing, logistics, retail, and other sectors with high automation needs. At the heart of its technology is the company's proprietary AI platform, KinetIQ, which enables the robots to understand their environment, plan complex tasks, and perform physical work in real-world industrial settings. The company pursues a fully integrated approach combining in-house hardware and AI development. The first beta systems are scheduled to go into productive use at customer sites as early as the fourth quarter of 2026. At the same time, Humanoid is preparing to launch mass production of its wheeled humanoid robots. Partnerships with industry leaders. Humanoid is already collaborating with several international industrial companies. Its partners include SAP, NVIDIA, Bosch, Siemens, and Schaeffler. The company particularly highlights a large-scale agreement with Schaeffler. The automotive and industrial supplier plans to deploy thousands of humanoid robots at its manufacturing sites and is also one of the investors in the current funding round. Bosch is also providing strategic support to Humanoid and will take over parts of the manufacturing process in the future. In addition, the company is contributing its expertise in hardware development, production, and supply chain management. Europe's response to the robotics boom. The market for humanoid robotics has so far been dominated primarily by companies from the U.S. and China. With this new funding, Humanoid is positioning itself as one of the key European players in the field of so-called Physical AI. The company now employs more than 250 engineers and researchers and maintains offices in London, Boston, Vancouver, and San Diego. With this funding, Humanoid aims to take the next step toward developing universally applicable robots that will work alongside humans in industry and logistics in the future. At the same time, the funding round sends a strong signal about Europe's ability to produce globally competitive AI and robotics companies.

Humanoid raises $152 million to expand robotics company. By PYMNTS | July 21, 2026 U.K.-based robotics firm Humanoid says it is now a $1 billion company. The startup achieved that status, becoming "Europe's first pure-play humanoid robotics unicorn," after raising $152 million in a Series A round announced Tuesday (July 21). "Now is a defining moment for the European technology ecosystem which positions the region among the global leaders in one of the world's fastest-growing tech sectors," Humanoid said in a news release. "Until now, much of the physical AI momentum has been concentrated in the United States and China. The investment demonstrates that the U.K. and continental Europe can produce and scale globally competitive humanoid robotics companies at the cutting-edge of the field." The company said the new capital will help it develop and launch its humanoid robotics platform, and help fund long-term commercial deployments with customers in sectors such as logistics, manufacturing and retail. In addition, the funds will help it begin mass manufacturing of wheel-based humanoid robots and to develop its proprietary AI brain KinetIQ. According to the release, Humanoid has secured partnerships with companies such as SAP, Nvidia, Bosch, and Siemens and recently inked "the industry's largest publicly announced commercial agreement" with Schaeffler to deploy thousands of humanoid robots for the manufacturing space. "Humanoid robotics will be one of the defining technologies of the next decade, reshaping how commercial and industrial work gets done," said Zia Huque, general partner at Prime Movers Lab, which led the round. "We expect the field to consolidate around a handful of category leaders across the US, Europe, and China. Humanoid AI will be one of a small cadre of robotics companies that will define humanoid robotics in Europe and beyond." In other robotics news, PYMNTS wrote recently about the rise of "cobots" - collaborative robots - which are designed to help workers with repetitive, physically demanding or precision-oriented tasks while letting their human counterparts tackle judgment-intensive activities. While this dramatically expands automation's potential, the report said, it also poses risk-related questions: Can autonomous systems function near human workers safely? Can these machines reliably interpret their environment? Can manufacturers ensure that intelligent systems will operate predictably under real-world conditions? "Those questions may determine the pace of industrial automation more than advances in robotics hardware itself," the report said. "In a traditional industrial setting, safety is achieved through distance. In a collaborative environment, safety must be engineered into every interaction. Robots must recognize people, anticipate movement, understand context and react appropriately to unexpected situations."