Notes from the
edge of the model.
Field notes on what actually breaks in production — agents, retrieval, evaluation, MLOps, and the career decisions nobody writes down. Longer arguments become papers; these are the rest.
353 posts
Let's have an honest conversation about AI and jobs.
Some jobs WILL be significantly impacted. Data entry. Basic content writing. Simple code generation. Routine customer service. Initial document review. Bookkeeping. Translation of…
By Pranay Mahendrakar Read postTesla Optimus. Figure 02. Boston Dynamics Atlas. Agility Digit. The humanoid robotics race is ON.
And the job market it's creating is unlike anything we've seen. Humanoid robots need: computer vision engineers who can handle real-time 3D perception. Reinforcement learning engineers…
By Pranay Mahendrakar Read postThe most interesting companies being founded right now aren't "adding AI" to existing products.
The difference matters enormously for job seekers. AI-native company: the product IS the AI. Every engineer touches ML. The culture is experimental. Shipping speed is everything. You'll…
By Pranay Mahendrakar Read postEvery major country is now building its own AI capabilities.
India: IndiaAI mission with billions in investment. Local LLMs for Indian languages. AI centers of excellence in Bengaluru, Hyderabad, Chennai, Pune. UAE: massive AI investment…
By Pranay Mahendrakar Read postThe real AI war isn't between OpenAI and Google.
Why this matters for engineers: the hardware you target shapes the software you write. NVIDIA dominates training today. CUDA is the default. But competition is intensifying fast. AMD's…
By Pranay Mahendrakar Read postHot take: we're approaching the end of the "LLM as the product" era and entering the "LLM as…
2023-2025: "We built a chatbot!" was impressive. 2026: "You built a chatbot?" is the baseline. The shift is from LLMs as the thing people interact with to LLMs as invisible…
By Pranay Mahendrakar Read postAlphaFold solved protein folding.
What's happening in 2026: AI-designed drugs entering clinical trials. Protein engineering with generative models. Genomic analysis at unprecedented scale. Synthetic biology guided by ML…
By Pranay Mahendrakar Read postI analyzed 200 remote AI job postings from the last quarter.
The good news: about 40% of AI roles are now remote-friendly, up from 25% two years ago. The talent shortage gives candidates leverage — companies that insist on office-only are losing…
By Pranay Mahendrakar Read postThe biggest leap in AI capabilities in 2026 isn't bigger models. It's reasoning models.
OpenAI's o-series. Anthropic's extended thinking. Google's reasoning improvements. These models don't just generate — they THINK before responding. Multi-step reasoning. Self-correction…
By Pranay Mahendrakar Read postLet me tell you which parts of the AI market are overheated and which are undervalued.
Overheated (too much money chasing too few real use cases): - General-purpose AI chatbot wrappers with no defensible moat - "AI-powered" versions of products that work fine without AI…
By Pranay Mahendrakar Read postThe next major platform shift won't be a new phone or computer. It'll be an AI operating system.
Think about it: instead of opening apps, you describe what you want done. The AI OS orchestrates multiple applications, services, and agents to accomplish the task. "Prepare my quarterly…
By Pranay Mahendrakar Read postAI is creating a category of jobs that didn't exist 2 years ago.
AI Data Trainer: Teaching AI systems to be better through feedback and evaluation. Companies like Scale AI, Anthropic, and OpenAI hire thousands. $25-$75/hr depending on expertise. AI…
By Pranay Mahendrakar Read post