AI FAQ

The average global smartphone replacement cycle has stretched to 3.5 years. Cameras, screens, and processors have reached a quality plateau where year-over-year improvements are incremental rather than transformative. Battery life has overtaken price as the top purchase driver for the first time, suggesting hardware differentiation has stalled.

Google gave Apple complete access to the Gemini model in Apple's own data centers. Apple uses a process called distillation, where smaller models learn from Gemini's reasoning outputs to produce efficient models with Gemini-like performance at a fraction of the compute. These distilled models can run on-device without an internet connection.

Apple's on-device Foundation Model is a roughly 3 billion parameter language model optimized for Apple Silicon through innovations like KV-cache sharing and 2-bit quantization. It runs at 30 tokens per second on iPhone 15 Pro and powers Apple Intelligence features including summarization, writing tools, and Siri enhancements.

Yes, and there are early signs of this. Samsung's Exynos 2600 markets 80 TOPS of NPU performance, more than double the prior generation. Samsung targets 800 million AI-enabled devices by end of 2026. But like megapixels before it, raw parameter count or TOPS may not correlate with actual user experience.

It depends on your current device. On-device AI requires specific hardware: Apple Intelligence needs an A17 Pro or later, and Android AI features require recent NPUs. If your phone is more than two generations old, you cannot run the latest on-device models at all. Morgan Stanley's 2026 survey found iPhone upgrade intentions at an all-time high of 37%, driven partly by AI capabilities.

Current smartphones run 1-3 billion parameter models natively. Apple's Foundation Model is roughly 3 billion parameters. Google's Gemini Nano ships at 1.8 to 3.25 billion parameters. Developers have also demonstrated running a 400 billion parameter Mixture of Experts model on iPhone 17 Pro, though only 17 billion parameters are active per inference pass.

IsoDDE (Isomorphic Drug Design Engine) is a unified computational drug design system released by Isomorphic Labs on February 10, 2026. It runs protein structure prediction, ligand binding, affinity estimation, and pocket identification in concert. On the hardest protein-ligand prediction tasks, IsoDDE achieves a 50% success rate versus roughly 23% for AlphaFold 3. On antibody-antigen modeling it exceeds AlphaFold 3 by 2.3× and Boltz-2 by 19.8×. On binding affinity prediction it achieves a Pearson correlation of 0.85, beating the physics-based gold standard FEP+ at 0.78.

No. As of February 2026, no AI-discovered drug has received FDA approval. The most advanced AI-discovered candidate is Insilico Medicine's rentosertib (ISM001-055), which reached Phase IIa with positive results published in Nature Medicine in June 2025. Isomorphic Labs targets its first clinical candidates for late 2026.

AI-discovered molecules show 80-90% Phase I success rates, well above the historical 40-65% average, indicating AI designs safe, tolerable molecules. However, Phase II success rates remain roughly 40%, the same as traditionally discovered drugs. AI has not yet demonstrated it can predict clinical efficacy, only safety and pharmacokinetic properties. If both trends hold, end-to-end success rates could roughly double from 5-10% to 9-18%.

Isomorphic has signed partnerships with Eli Lilly ($45M upfront, $1.7B in milestones), Novartis ($37.5M upfront, $1.2B+ in milestones, expanded February 2025), and Johnson & Johnson (terms undisclosed, announced January 2026). Total disclosed value exceeds $4 billion, but the 50:1 ratio between milestone promises and actual upfront cash reflects pharma's caution about AI drug discovery outcomes.

Isomorphic leads on computational benchmarks but trails on clinical progress. Insilico Medicine has 10+ IND approvals across 31 programs and the most advanced clinical candidate (rentosertib, Phase IIa). Recursion, which absorbed Exscientia in a $688M merger, takes a phenomics-first approach with 65 petabytes of biological imaging data. Both own wet-lab infrastructure that Isomorphic lacks. Isomorphic's advantages are Alphabet-scale compute, the AlphaFold lineage, and its unified architecture.

Isomorphic sits within Alphabet's Other Bets segment, which posted a $3.6 billion operating loss in 2025 against Alphabet's $132 billion net income. Isomorphic raised a $600 million external round led by Thrive Capital in March 2025. Alphabet can sustain this investment indefinitely, which is itself a competitive advantage over cash-constrained biotech startups.

Isomorphic Labs CEO Demis Hassabis said at Davos 2026 that the company expects its first AI-designed drugs to enter clinical trials by the end of 2026, pushing back an earlier target of 2025. The company hired Dr. Ben Wolf as CMO in June 2025 and opened a Cambridge, Massachusetts office, signaling clinical-stage staffing.

Several recent proofs demonstrate structural constraints. Duman Keles et al. (2023) proved O(n²) attention complexity is a necessary lower bound. Kalai and Vempala (STOC 2024) proved any calibrated language model must hallucinate at a certain rate. Chowdhury (2026) showed the lost-in-the-middle problem is geometric, present at initialization before training. These are not engineering challenges to be fixed with better data.

Not a single replacement but a hybrid stack. Over 60% of frontier models already use Mixture of Experts. Production systems like AI21's Jamba, Alibaba's Qwen3-Next, and Microsoft's Phi-4-mini-flash-reasoning blend attention with state space models (Mamba) for 3-10x throughput gains. Diffusion language models like LLaDA offer a wilder alternative, generating text through denoising rather than sequential token prediction.

It is already happening. DeepMind's AlphaEvolve found a 23% kernel speedup inside Gemini's own architecture. Karpathy's AutoResearch discovered about 20 improvements on his own highly-tuned codebase, cutting the metric by 11%. Sakana AI's AI Scientist v2 produced the first AI-authored paper accepted through standard peer review. The timeline from thought experiment to working systems was faster than most expected.

For dense transformers, evidence points to flattening. OpenAI's Orion model hit GPT-4 performance after just 20% of training, with diminishing returns for the remaining 80%. But test-time compute opened a different axis: inference spending hit $2.3 billion at OpenAI in 2024, 15x training costs. The Densing Law shows capability per parameter doubling every 3.5 months through MoE, distillation, and better data curation.

MCP is agent-to-tool; A2A is agent-to-agent. MCP connects AI agents to tools and data sources. A2A connects agents to each other for multi-agent collaboration. They operate at different architectural layers and are complementary, not competing.

Start with MCP, then add A2A. For most enterprise deployments, MCP handles tool integration first, then A2A layers on when you need multi-agent coordination across organizational boundaries. AWS, Microsoft, Salesforce, SAP, and IBM already support both protocols.

Both are under the Linux Foundation. MCP sits within the Agentic AI Foundation (AAIF), which has 146 member organizations including Anthropic, OpenAI, and Block. A2A has its own governance body with 150+ partner organizations.