AI for Rapid Antibody Discovery🧬🔬

Chai-2 achieves 16% binding in zero-shot antibody design; Genies debuts smart AI avatars for immersive games; Apple plans AI-powered glasses, Vision Air, and next-gen XR by 2030.

This Week in Tech & Innovation: AI, Avatars, and Apple’s AR Ambitions

Welcome to this week’s edition of The Future Lens, where we spotlight the most groundbreaking advancements shaping technology, science, and creativity.

🔬 AI Revolution in Antibody Discovery:
Biotech takes a massive leap forward with Chai-2, a next-generation generative AI model by Chai Discovery. Redefining how antibodies are designed, Chai-2 achieves 100x performance gains over traditional methods—compressing development timelines from months to just two weeks. With a 16% binding success rate across 52 novel targets, this zero-shot design tool is paving the way for faster, programmable drug discovery at atomic precision.

🎮 The Rise of Smart Avatars & AI NPCs:
Genies is transforming digital experiences with emotionally intelligent AI avatars. These avatars can speak, gesture, evolve, and scale across games and platforms, enabling developers to create rich, immersive universes with minimal effort. Coupled with a powerful UGC ecosystem and revenue-sharing model, Genies is setting a new standard for interactive storytelling and AI-powered gameplay.

👓 Apple’s AR & Smart Glasses Roadmap (2025–2030):
Apple is charting a long-term vision in spatial computing. While 2025 brings only an M5 chip upgrade to the Vision Pro, the real excitement begins in 2027 with AI-powered smart glasses—no display, but full of voice and gesture interaction. The decade will see launches of Vision Air, second-gen Vision Pro, and XR Glasses with color displays, all backed by Apple’s hardware prowess and ecosystem integration. The challenge? Building truly compelling AI-driven software to match.

Let’s dive deeper into each of these stories below. 👇

Chai-2: Revolutionizing Antibody Design with AI Precision

Chai Discovery has announced Chai-2, a groundbreaking generative AI model for zero-shot de novo antibody design, achieving double-digit hit rates and outperforming previous computational methods by 100-fold. Chai-2 enables rapid, precise, and programmable antibody discovery, compressing timelines from months or years to just two weeks by routinely finding viable binders with only 20 designs per target. In tests across 52 novel protein targets, Chai-2 achieved a 16% binding rate and found at least one successful binder for half the targets, with even higher success in miniprotein binder design. The model’s multimodal architecture integrates atomic-level structure prediction and generative modeling, creating novel, diverse, and epitope-specific binders across various biologic formats. Chai-2’s capabilities promise to revolutionize molecular engineering by making computational-first, deterministic design central to drug discovery, with early access now open for academic and industry partners focused on impactful biomedical challenges.

Genies AI: Powering Next-Gen Smart Avatars and Immersive Gameplay

Genies’ AI Avatars are advanced, visually expressive, and emotionally responsive digital characters designed for seamless integration into games and digital experiences, enabling developers to rapidly create customizable, dynamic Smart Avatars and AI NPCs at scale514. With Genies’ tool suite, developers can easily generate game-ready avatars, assets, and environments, streamline asset creation through auto-rigging and optimization, and support user-generated content (UGC) for new gameplay genres12. These avatars, powered by large language models and behavioral AI, can engage in natural conversations, adapt to user behavior, and animate with lifelike gestures—making them persistent, evolving digital identities that can move across platforms and experiences124. Genies’ technology also supports revenue opportunities for creators via Party Platforms, where users can co-create games and digital goods with iconic IP, publish directly to built-in audiences, and participate in revenue sharing.

Apple Vision & Glasses: 2025–2030 Plans

Apple is advancing its Vision and smart glasses lineup with a detailed roadmap through the end of the decade, according to analyst Ming-Chi Kuo. In 2025, Apple will refresh the Vision Pro with an M5 chip but no major changes, maintaining its niche status. No new hardware is planned for 2026. In 2027, Apple will launch its first smart glasses, similar to Meta’s Ray-Bans, offering voice control, gesture recognition, and AI-powered environmental awareness, but without a display; these glasses are expected to reach significant sales volumes. That year will also see the debut of the lighter, more affordable Vision Air headset. In 2028, Apple plans to release a true second-generation Vision Pro with a new design, lighter weight, Mac-grade processor, and lower price, alongside advanced XR Glasses featuring a color display and waveguide optics. Looking beyond 2028, Apple is developing a second-generation XR Glasses variant and has paused work on a display accessory while reviewing its competitive positioning. Apple’s strategy leverages its hardware strengths and ecosystem integration, but faces challenges in developing compelling AI-powered features and software.

Hand Picked Video

In this video, we’ll look at how an AI Nutrition Agent creates personalized meal plans based on your health data, dietary preferences, and goals—automating research to optimize your diet effortlessly!

Top AI Products from this week

  • Portia AI - Portia enables your developers to build safe, reliable AI agents in industries where compliance comes first. Create human readable plans for your agents, store where the agent is in its execution and define guardrails for when the agent should pause execution to authorise with a person or run additional checks.

  • String.com - Prompt, run, edit, and deploy AI agents in seconds.

  • Lazy - Context switching is poison for productivity. When you get that brilliant idea or see something inspiring, you should be able to take smart notes without switching app or tabs. That’s Lazy. One ⌘ shortcut to capture anywhere. Sign up to our waitlist at lazy.so!

  • Unify - Unify is a system of action that helps you reach the right customer, at the right time, with the right message, at scale. The platform enables this through intelligent prospecting based on buyer signals, hyper-personalized messaging to audiences.

  • UniDeck - UniDeck makes it easy to create personalized dashboards and portals for both individuals and enterprises. Connect your favorite tools like Jira, Google, Microsoft Teams, Trello, and GitHub, while AI helps you design and automate effortlessly.

  • zookish - An embeddable conversational AI to make your website talk and turn your website visitors into leads. Build automations using natural language and MCPs, start now with our generous Free Access.

This week in AI

  • AB-MCTS AI Collective Intelligence - Sakana AI’s AB-MCTS lets multiple frontier LLMs collaborate at inference, boosting ARC-AGI-2 benchmark scores via adaptive trial-and-error and model selection.

  • Cloudflare No AI Crawl Without Pay - Cloudflare now blocks AI crawlers by default unless they compensate content creators, marking July 1 as "Content Independence Day" to ensure fair value for original content.

  • X Pilots AI-Generated Community Notes - X is testing AI chatbots, including Grok and third-party bots, to draft Community Notes. All AI notes require human review and consensus before public release.

  • Grammarly Acquires Superhuman for AI Suite - Grammarly is acquiring email startup Superhuman to boost its AI productivity platform; Superhuman CEO and team will join, expanding AI tools for email and workflow.

  • DeepFleet AI Boosts Robot Fleet Efficiency - Amazon’s new DeepFleet AI foundation model optimizes robot routes, cutting travel time by 10% and enabling faster, more efficient deliveries across its global network.