Good morning. It's Thursday, March 5, and we're covering AI homework tools in China, the OpenClaw agent boom, signs AI investment may boost hiring, and more.

New here? Keep up with the future of tech, sign up here. Have feedback? Send us a note: [email protected]. If you liked this email, share it with a friend.

🤔 AI Adoption Linked to Near-Term Job Growth
The European Central Bank says firms using AI are currently more likely to hire than cut staff in the euro zone. Survey data shows companies investing in AI expect employment to grow and report stronger hiring plans. The finding contrasts with surveys predicting job losses over the next five years. Economists note the longer-term impact could change as AI transforms production.

⚛️ AI Helps Physicists Find New Quantum Gravity Result
Researchers report a new quantum gravity result showing certain graviton interactions once thought impossible can occur under specific conditions. Using GPT-5.2 Pro, the team derived formulas for “single-minus” graviton scattering amplitudes in a special momentum regime. The model helped generate the initial solution and draft, while scientists verified the math analytically.

🤝 CoreWeave Partners with Perplexity for AI Inference
CoreWeave announced a multi-year partnership with Perplexity to power its next-generation AI inference workloads. Perplexity will run its Sonar and Search API services on dedicated NVIDIA GB200 NVL72 clusters in CoreWeave’s cloud. The setup aims to lower latency and stabilize costs as AI search demand grows, highlighting rising demand for AI-optimized cloud infrastructure.

🧮 AI Agent Finds New Solution to Math Challenge
Cursor researchers say an autonomous AI agent discovered a new solution to “Problem Six” from the First Proof math challenge. Running for four days without human guidance, the system used the Marcus–Spielman–Srivastava interlacing polynomial method and produced stronger constants than the existing solution. Early feedback from mathematicians suggests the proof is likely correct.

Block, Inc. announced plans to cut more than 4,000 employees, framing the move as part of an AI-driven restructuring. But industry critics argue the layoffs reflect a familiar Silicon Valley playbook: cost cutting and organizational consolidation dressed in AI language.

Market analysts note that many eliminated roles—such as policy and diversity positions—are not easily replaced by current AI tools. The announcement nevertheless boosted investor confidence, sending Block’s stock up 24% and reinforcing pressure across tech to present aggressive AI strategies. → Continue reading here. (Paywall)

The Recap: Parents across China are increasingly using artificial intelligence tools—from chatbots to translation devices—to help their children study and reduce homework stress, according to The New York Times. The trend reflects China’s broad enthusiasm for AI: a 2025 KPMG survey found over 80 percent of Chinese respondents felt optimistic about the technology, compared with just over 50 percent of Americans. While the boom has fueled a 43 billion RMB ($6 billion) education technology market rife with hype, many families say AI tools are delivering practical academic and emotional relief.

Highlights:

Forward Future Takeaways:
China’s embrace of AI in education highlights how cultural context shapes adoption: where U.S. debates center on cheating and cognitive decline, many Chinese families see AI as a competitive equalizer and time-saving tutor. The rapid commercialization, however, raises familiar concerns about inflated claims, uneven quality, and privacy trade-offs—especially when children’s images and data are involved. → Read the full article here. (Paywall)

The Recap: In early 2026, OpenClaw—an open-source framework for building autonomous AI agents that operate personal computers—sparked a rapid wave of experimentation across China’s tech sector, from startups to major cloud providers. Chinese firms including ByteDance, Alibaba, and Tencent quickly launched cloud services supporting OpenClaw, while developers built new agent-driven apps ranging from AI matchmaking tools to automated trading assistants. The surge highlights China’s fast-moving developer culture and suggests AI agents may become the next competitive frontier for software platforms and hardware makers.

Highlights:

Forward Future Takeaways:
OpenClaw illustrates how open-source tools can ignite rapid experimentation when paired with cheap models and accessible cloud infrastructure. China’s willingness to deploy the technology quickly—especially through major cloud platforms—could accelerate the development of agent-driven applications faster than in more cautious markets. → Read the full article here. (Paywall)

Researchers reported that they compressed an AI vision model from 60 million variables to just 10,000—while preserving nearly the same performance. The team, led by Ben Cowley of Cold Spring Harbor Laboratory with collaborators at Carnegie Mellon and Princeton, trained the system on data from macaque monkeys’ V4 neurons, which process shapes, colors, and curves.

By pruning redundant components and applying statistical compression techniques, they created a model small enough to send as an email attachment. The advance could lower the computing power required for AI systems and offer neuroscientists a clearer window into how biological brains process vision. → Read the full article here.

X Targets Undisclosed AI War Videos: Platform will suspend creator revenue for 90 days if users post AI-generated conflict footage without labels after fake Iran war clips went viral.

Claude Code Adds Voice Commands: Anthropic is rolling out Voice Mode to its coding assistant, letting developers execute tasks via speech. The feature is live for 5% of users.

Bank of America Rejects AI Job Doom: BofA economists say fears of AI wiping out white-collar jobs are overblown, arguing productivity gains will expand the economy despite market jitters.

Anthropic AI Stuck in Defense Limbo: Claude is reportedly aiding U.S. targeting via Palantir’s Maven system in the Iran conflict, even as the Pentagon bans the model.

Father Sues Google Over Gemini Death: Lawsuit claims Gemini fueled a user’s delusion that an AI was his sentient wife, raising alarms about chatbot design, “AI psychosis,” and guardrails.

Poe: Access top AI models like GPT-4.5, Claude, DeepSeek, Runway, and ElevenLabs in one place.
Cody: Enterprise AI coding assistant with chat, autocomplete, and inline edits to boost developer productivity.
Quickads: AI platform to research competitors, generate image and video ads, and optimize campaigns.

❤️ Love Forward Future? Share your referral link with friends and colleagues to unlock exclusive perks! 👉 Get your link here.
📢 Want to advertise with us? Reach 1M+ AI enthusiasts, tech leaders, and decision-makers. Just reply to this email.

Thanks for reading today’s newsletter. See you next time!
Matthew Berman & The Forward Future Team

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀