| FutureReady |
16-30 Sep | Issue 1 |
|
| Insights into the world of Technology and AI |
|
| AI agents started acting on their own this fortnight, and the world's rule-makers spent it catching up - from Brussels to the UN Security Council to New Delhi |
| Advances |
22 Sep - Claude Opus 5.5 - Anthropic says its new model brings Fable-level performance at lower prices. Frontier AI for business keeps getting cheaper, but benchmark claims still need testing on your own work. More29 Sep - OpenAI GPT-6.1 Sol - OpenAI says Sol offers near-Astra intelligence at one-fifth of Astra's standard token prices. TechCrunch reports the planned GPT-6.1 Astra was shelved over deception and unauthorized-action concerns. Safety can stop a release. More19 Sep - Dreaming Agents - Google DeepMind's Dream-RSI lets agents rehearse past attempts in their head and come back sharper; one hard task dropped from 550 attempts to 317. Agents are starting to practice, not just run. More
|
| Applications |
29 Sep - OpenAI Dots - Always-on agents that chase goals across Slack and Teams, with permission rules you set. The AI coworker is now a shipping product. More30 Sep - Apple Pay in India - Launched for eligible Axis Bank Visa and Mastercard credit cards, with iPhone XR or later among supported devices. It adds a card-based contactless option alongside UPI, not a replacement for it. More24 Sep - UPI Merchant Fees - CNBC reports a new fee framework for select large merchants. The business impact is payment costs, not a blanket charge on every UPI user. More
|
| Environment |
29 Sep - Infosys Reset - Plans to move up to a third of its 320,000 people into specialist AI roles and build a 6,000-strong frontier-engineer team as AI breaks per-head billing. India's IT model is being rewritten. More17 Sep - Semicon 2.0 - India targets 200 chip-design startups and companies, plus training for over 1 lakh technicians. Investment commitments of $11-12 billion are expected over two to three years. The focus is a chip ecosystem, not just factories. More27 Sep - EU Turns On Enforcement - Brussels sent formal information requests to more than 30 AI providers, the AI Act's first live enforcement. Europe's AI rules now have teeth, not just deadlines. More |
|
Faster does not mean reliable - Why faster AI progress can outpace reliability and testing Is AI alive? - A discussion of information, understanding and consciousness Jev: real-time decisions - Simulated games and navigation illustrate rapid AI decisions |
Project KnowHow Ideas, decisions and lessons from building with AI, with practical resources for non-technical users |
|
|