SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World
Back to feed
Programming

Google Rolls Out Gemini 3.7 Flash, Transcription Model and Live Upgrades in August 2026

0
·1 views

Google released a broad wave of Gemini updates in August 2026, spanning new models, the Gemini app, and Gemini Live. Gemini 3.7 Flash, aimed at coding and agent workflows, is priced at half the cost of Gemini 3.6 Flash through December 31, 2026, and is being rolled out to Google AI Pro and Ultra users via Gemini Spark. Late in August, Google also launched Gemini 3.5 Transcribe, an enhanced speech-to-text model designed to convert spoken content from calls, meetings, and recordings into searchable text. Gemini Omni 1.1 Flash, which adds advanced video generation, is available across Google Flow, AI Studio, the Gemini Enterprise Agent Platform, and the Gemini app. Gemini Live received proactive features including Personal Intelligence, Daily Brief, and hands-free inbox management, signaling Google's push to embed Gemini into continuous workflows rather than single-session chat interactions.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Developer builds AI agent to play Slay the Spire 2 using rejection-feedback, not LLM calls

A developer built an autonomous agent that completed a full solo combat session in Slay the Spire 2 on August 9, 2026, executing 35 actions across 7 turns without any human input. Notably, none of the decisions were made by a large language model; instead, a local score-based greedy policy running inside a Python bridge handled all in-combat choices. The project was enabled by reverse-engineering the game's internal C# codebase, where the developer discovered an existing AutoSlay system and a public card-selection interface called ICardSelector. A key engineering insight was the rejection-feedback pattern, where game errors became persistent state used to filter the agent's next actions — requiring no retraining or re-prompting. The developer also navigated two major technical failures: a main-thread deadlock in Godot and brittle patches that broke with each Early Access update.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

React 18's useSyncExternalStore Offers a Cleaner Way to Read RxJS BehaviorSubjects

React 18 introduced useSyncExternalStore, a built-in hook that provides a straightforward way to connect RxJS BehaviorSubjects to React components without third-party libraries or manual useEffect subscriptions. The hook accepts a subscribe function and a getSnapshot function, mapping directly onto a BehaviorSubject's existing subscribe and getValue methods. Developers can build a reusable adapter hook by wrapping these two functions with useCallback to prevent unnecessary resubscriptions on every render. A key pitfall is defining the subscribe function inline without memoization, which causes React to tear down and recreate the RxJS subscription on every render, risking missed state updates. The same primitive is used internally by state management libraries like Redux, Zustand, and Jotai, making it a reliable foundation for external store integrations.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

OpenAI Claims AI Swarm of 10,000 Agents Cracked Major Unsolved Math Problem

OpenAI published a paper on September 8, 2026, claiming an unreleased AI system produced a proof for the Navier–Stokes existence and smoothness problem, one of seven Millennium Prize Problems carrying a $1 million reward since 2000. The system reportedly used a coordinated swarm of roughly 10,000 AI agents running in parallel, with different groups tackling the problem from multiple angles simultaneously. Researchers periodically extracted promising partial results and fed them back into the system to cross-pollinate ideas across separate reasoning threads. The entire effort spanned approximately 88 hours of agent work, followed by 17 additional hours to formalize and verify the proof using Lean, a mathematical proof-verification language. However, the paper comes with significant caveats, as the underlying AI model remains unreleased and independent expert scrutiny of the result is still ongoing.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Why Most Agile Transformations Change the Language but Not the Power Structure

A widely cited pattern in corporate Agile transformations shows that organizations adopt Agile ceremonies and vocabulary while leaving core decision-making authority untouched. Digital transformation spending in Western markets was projected at $6.3 trillion between 2022 and 2024, yet failure rates remain around 84%. Middle managers, whose roles are most directly threatened by genuine Agile adoption, often preserve existing mechanisms by renaming them rather than restructuring them. Funding, headcount, and strategic direction continue to be controlled by the same leadership layers that existed before the transformation began. Critics argue that many such initiatives are designed not to redistribute decision-making power, but to create the appearance of change while keeping organizational hierarchies intact.

0 comments Read more at DEV Community