Good morning. It's Friday, February 13, and we're covering AI fundraising, Gemini copycat attacks, AI-driven coding shifts at Spotify, 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.

🧑‍🤝‍🧑 Simile raises $100M to build AI “digital twins”
Simile, an AI startup that creates digital twins of individuals to simulate behavior, has raised $100 million in funding led by Index Ventures. Backers include Bain Capital Ventures, AI researcher Fei-Fei Li and OpenAI co-founder Andrej Karpathy. Simile’s model uses interview data, transaction logs and scientific texts to predict how people might respond to products, UI changes or even earnings calls.

👨‍💻 Spotify Engineers Stop Writing Code
Spotify said its top developers haven’t written code since December, relying instead on AI tools like Claude via its internal system “Honk.” Co-CEO Gustav Söderström said engineers can deploy fixes or features from Slack on their phones before arriving at work. Spotify shipped 50+ updates in 2025 and rolled out new AI features, saying its proprietary music data gives it a defensible edge in training.

💰 Anthropic raises $30B at $380B valuation
Anthropic announced that it had raised $30 billion in funding at a $380 billion post-money valuation. The company said the capital will be used to deepen research and continue innovating its AI systems. The announcement underscores the ongoing investment boom in large-scale AI development and positions Anthropic alongside the most heavily funded players in the field.

🔮 Musk Predicts End Of Coding
Elon Musk believes programming will “die” this year as AI begins generating optimized machine code directly. The argument suggests future systems will translate human intent into executable software without traditional languages or compilation.It also suggests pairing this with Neuralink for “imagination-to-software” creation, framing AI as replacing coding entirely — though the idea remains speculative.

It feels responsible. Productive. Almost virtuous. Swipe away every app and your phone should run smoother… right? Stick around for the answer! 👇️

👉 Watch on X or YouTube.

Dario Amodei told Ross Douthat on the Interesting Times that advanced AI could deliver “a country of geniuses in a data center” within a few years—accelerating medical breakthroughs and pushing developed-world GDP growth into the 10–15% range. But he warned the same systems could rapidly displace entry-level white-collar jobs, destabilize legal and political institutions, and intensify U.S.–China competition.

Amodei argued that AI misuse—such as autonomous drone swarms or biological weapon design—poses serious risks, and said meaningful international restraint would require enforceable verification. He also acknowledged deep uncertainty about AI consciousness, saying “we don’t know if the models are conscious,” while describing internal safeguards like a 75-page “constitution” guiding Anthropic’s models. → Read the full article here.

Google said its AI chatbot Gemini faced “distillation” or model-extraction attacks, including one campaign that sent more than 100,000 prompts to probe its inner workings. The company described the activity as commercially motivated efforts by private firms or researchers seeking to replicate Gemini’s capabilities.

Google’s Threat Intelligence Group said such attacks are likely to become more common as companies deploy custom large language models (LLMs) trained on sensitive data. Google considers the practice intellectual property theft and has adjusted Gemini’s defenses after detecting the activity. → Read the full article here.

A recent arXiv paper reports that a scaffolded version of Gemini produced novel mathematical contributions by connecting tools from geometric analysis to approximation algorithms—an approach the authors say has rarely been used in that context. In one example, the model linked the Kirszbraun Extension Theorem to Steiner tree computation in what researchers describe as a new and “natural” connection.

The result suggests that structured prompting and iterative reasoning scaffolds may enable large language models to generate non-trivial cross-field insights. While the work does not demonstrate broad scientific autonomy, it indicates that tooling and orchestration—rather than raw model scale alone—may be key to unlocking incremental, domain-specific breakthroughs. → Read the full article here.

🤝 Anthropic Funds AI Safety PAC (Paywall): Anthropic commits $20M to a super PAC backing stricter AI rules, setting up a midterm clash with OpenAI-aligned groups that have raised over $50M.

🐌 Apple’s Siri Upgrade Delayed (Paywall): Apple’s planned Siri overhaul hit testing snags, likely delaying key features from iOS 26.4 in March to later updates, potentially iOS 26.5 in May.

✨ Google Upgrades Gemini Deep Think: Google launches an enhanced Gemini 3 Deep Think mode that turns sketches into 3D-printable files, rolling out to AI Ultra subscribers.

🦾 OpenAI Taps Cerebras for Speed: OpenAI launches GPT-5.3-Codex-Spark on Cerebras chips, claiming 15x faster code generation in its first major move beyond NVIDIA amid strained ties.

🚀 MiniMax Releases M2.5 Model: Shanghai-based MiniMax unveils open-source M2.5, touting 80.2% on SWE-Bench Verified and speeds up to 100 tokens/s, offering low-cost APIs from $1 per hour.

Closing background apps usually does not save battery life — and can actually use more. When you leave an app on modern iOS or Android, it enters a suspended state. It’s frozen in RAM, not actively using the CPU. Reopening it simply “unfreezes” it, which takes minimal energy.

Force-closing is different. When you swipe an app away, the system has to fully reload and reinitialize it the next time you open it. That extra CPU work consumes more power than resuming from suspension. Both Apple and Google have confirmed that routinely force-closing apps isn’t necessary — the operating system already manages memory efficiently.

❤️ Love Forward Future? Share your referral link with friends and colleagues to unlock exclusive perks! 👉 Get your link here.

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

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀