Google just rolled out Studio Bot, a dedicated coding assistant built directly into Android Studio to generate code, fix bugs, and answer dev questions. It runs on the new Codey model derived from PaLM 2, supporting Kotlin and Java. It is currently limited to US developers on the Canary channel.
Google just unveiled its next-gen PaLM 2 language model at I/O, powering the newly updated Bard and serving as the foundation for the company's latest AI updates. Instead of chasing pure parameter scale, PaLM 2 sharpens logic, math, and support for over 100 languages. It also spawns specialized variants like Codey and Med-PaLM 2, plus a mobile version hitting 20 tokens per second.
Airtable just introduced Airtable AI, bringing generative AI components directly into its no-code enterprise platform. The update adds a new 'AI field type' and pre-built templates for tasks like marketing plans and job specs without writing code. It is currently open for early access as a paid add-on, with full launch slated in about three months.
IBM just launched Watsonx, a comprehensive enterprise studio designed to help companies train, tune, and deploy custom generative AI models. The suite bundles an AI studio, a governance layer, and specialized foundation models alongside a partnership with Hugging Face. Why it matters: IBM is pitching enterprise trust and private cloud deployment to win over businesses still wary of Big Tech rivals.
Hugging Face and ServiceNow just released StarCoder, an open 15-billion-parameter code-generating model. Trained on over 80 programming languages using permissively licensed data, it allows royalty-free use even for enterprises. Why it matters: It delivers a transparent, customizable alternative to proprietary tools like GitHub Copilot without the subscription fee.
Digital artist Rhett Dashwood just turned a $69 budget and zero coding experience into a $50M meme coin called $TURBO using GPT-4. The AI guided the tokenomics, drafted the white paper, and wrote the Solidity smart contracts, while Midjourney generated the mascot. It serves as a wild proof of concept for non-technical creators building projects with AI.
NVIDIA just open-sourced NeMo Guardrails, a software suite designed to keep LLM-powered applications accurate, safe, and on-topic. It adds topical, safety, and security guardrails to models like ChatGPT, integrating with tools like LangChain and Zapier via just a few lines of code. Why it matters: developers don't need to be ML experts to prevent their chatbots from going rogue.
Google just updated Bard with full programming support, tackling one of its most-requested features. The experimental AI can now generate, debug, and explain code across more than 20 languages like Python, C++, and Go, plus export directly to Google Colab. Google notes it's still experimental and prone to bugs, so you will definitely want to double-check its output.
Cohere just released The Embedding Archives, open-sourcing around 100 million multilingual embedding vectors across 10 languages from Wikipedia. Available on Hugging Face, the 768-dimensional vectors also come with a 10-million subset hosted on Weaviate for direct semantic search. Why it matters: Developers can skip the slow, costly process of text chunking and vectorization when building search engines.
Webflow just announced it is integrating generative AI directly into the Webflow Designer. Users will soon be able to generate layout components, rewrite CMS copy, auto-generate SEO settings, and customize templates using natural language prompts. Why it matters: bringing AI inside a visual editor solves the headache of manually stitching raw AI-generated code into working websites.
Amazon just launched Bedrock, a serverless cloud platform letting builders access top-tier AI models from Anthropic, Stability AI, and its own Titan family via API. Developers can privately fine-tune models with as few as 20 labeled examples without managing clusters. Why it matters: AWS is turning into an enterprise AI bazaar, embedding generative tech straight into its existing cloud stack.
Google just partnered with cloud coding platform Replit to bring its foundation AI models directly into Replit's Ghostwriter assistant. Coming just a week after GitHub unveiled Copilot X powered by GPT-4, the alliance sets up a fierce battle between Google and Microsoft over who will power the next generation of software developers.
Ubisoft just introduced Ghostwriter, an in-house AI tool designed to generate first drafts of NPC chatter and combat barks. Writers select and edit variations through pairwise comparisons, letting the model learn preferred phrasing over time. While meant to save writers from repetitive grunt work, the announcement immediately drew pushback from developers urging studios to keep hiring humans.
GitHub just unveiled Copilot X, a major vision powered by OpenAI's GPT-4 that moves AI far beyond basic code completion. It integrates a ChatGPT-style chat directly into IDEs, automatically drafts pull request descriptions, answers questions on documentation, and assists in the CLI. Basically, your AI pair programmer is stepping up to cover the full software development lifecycle.
The PyTorch team just rolled out PyTorch 2.0, introducing a major compiler revamp that boosts performance without breaking backward compatibility. The marquee feature, torch.compile, leans on OpenAI Triton to deliver an average 36% speedup under AMP precision across 165 models. Native scaled dot product attention is also baked in, eliminating the need for external FlashAttention packages.
Google just unveiled a sweeping wave of generative AI updates across its ecosystem. Google Workspace users will soon be able to draft emails and documents simply by typing a prompt, while developers get access to the new PaLM API and MakerSuite to prototype custom AI apps. Vertex AI also gains foundation model support for enterprise workloads. It's Google's direct answer to Microsoft's rapid AI push into productivity software.
OpenAI just made GPT-3.5 Turbo and Whisper available via API, slashing ChatGPT costs by 90% down to $0.002 per 1,000 tokens. Whisper speech-to-text runs at $0.006 per minute, and OpenAI will no longer train on developers' API data by default.
Andreessen Horowitz just announced that it is leading Replicate's Series A round. Founded by the creator of Docker Compose and a former Spotify machine learning engineer, the platform aims to let developers run and fine-tune open-source models with clean abstractions instead of messy CUDA errors. Why it matters: bridging the gap between 30 million software engineers and scarce machine learning specialists is how AI software actually goes mainstream.
Roblox is testing two new generative AI tools for Roblox Studio that turn natural language prompts into 3D materials and working code. The goal is to let creators jump straight into world-building even if they lack 3D modeling skills or coding experience. Both features will roll out to Roblox Studio developers in the coming weeks.
OpenAI just recruited roughly 1,000 remote contractors in regions like Latin America and Eastern Europe, with about 40% tasked with writing step-by-step coding explanations to train its models. Rather than scraping raw GitHub code, the company is capturing human reasoning and debugging workflows. Why it matters: Teaching models human thought processes could soon automate entry-level software development entirely.
