Gemini 3.5
AdoptPlatforms
Google's Gemini family of AI models used for multimodal and conversational tasks.
Why it's here
Placed in Adopt: 24 article(s) of evidence from 7 source(s), led by research-stage coverage, with 8 in the last 30 days. Confidence 100%.
Evidence (13)
- 5Hacker News·7/25/2026framework_updateEngineering management after AI lowered the cost of code
The article argues that the drop in code-generation cost from LLMs should change management practices only where those practices depended on writing code being expensive. It cautions that claims of major productivity gains are not yet proven and that organizations should measure real business outcomes and system health rather than relying on old velocity-style proxies.
- 4Simon Willison·7/22/2026researchBenchmarking Whether AI Labs Are "Pelicanmaxxing"
An analysis tested whether major AI labs may be optimizing models for Simon Willison’s informal “pelican riding a bicycle” benchmark. The author generated 1,008 SVGs across seven frontier models, scored them with LLM judges, and found no clear evidence that pelican-on-bicycle outputs were better than other animal-vehicle combinations. The article provides the full experiment and code on GitHub.
- 7Ars Technica AI·7/21/2026model_releaseGoogle launches Gemini 3.6 Flash and a cybersecurity model
Google introduced three new AI models, including Gemini 3.6 Flash and its first Gemini model aimed at cybersecurity. Gemini 3.5 Flash has been deprecated, while the delayed Gemini 3.5 Pro is still not available and Gemini 4 was only teased. The company says 3.6 Flash improves coding and multimodal performance based on user feedback.
- 9Ars Technica AI·7/16/2026regulationEU orders Google to open Android AI access and share search data
The European Commission has issued binding DMA measures that require Google to improve interoperability on Android and provide broader access to search data. Google says the changes could weaken privacy and security, but it must comply as a designated gatekeeper under the law.
- 4Hacker News·7/16/2026product_launchDecoy Font hides text from AI readers
Decoy Font is a TTF typeface that uses hybrid-image style spatial frequencies to show different text to humans and AI/OCR systems depending on viewing distance. The project is positioned as an anti-scraping and anti-OCR tool, but it is not a guaranteed defense against more capable agents.
- 4Hacker News·7/16/2026product_launchNotebookLM renamed to Gemini Notebook
Google has renamed NotebookLM to Gemini Notebook as part of its ongoing Gemini branding. The product remains an AI-powered note-taking and research tool, with the announcement shared on Google's blog and discussed on Hacker News.
- 6Hacker News·7/15/2026regulationWhy I Left Google DeepMind
The author describes efforts to pressure Google to stop serving U.S. immigration enforcement agencies and to resist a Pentagon military AI deal. The piece argues that internal appeals, proposals, and public lobbying failed to prevent what the author sees as unethical government contracting, leading to the decision to leave Google DeepMind.
- 3Hacker News·7/14/2026researchAre We Outsourcing Too Much Thinking to AI?
This essay reflects on how people increasingly use AI to handle decisions, research, and reasoning that used to require active human judgment. It contrasts AI assistants with search engines, arguing that modern systems can reduce the need to think through problems step by step and may erode personal autonomy if overused.
- 2Hacker News·7/9/2026researchDeveloper says LLM output fatigue is setting in
A developer describes spending hours each day using LLMs for coding, search, and code review, and says the tools have made them more productive. Over time, however, repeated exposure to similar writing patterns, hallucinations, and stylistic quirks has become tiring and frustrating.
- 7Google DeepMind·6/30/2026model_releaseGoogle launches Nano Banana 2 Lite for faster, cheaper image generation
Google DeepMind has introduced Nano Banana 2 Lite, a new image-generation model positioned as the fastest and cheapest option in its lineup. It is part of the Gemini 3.1 family and is aimed at rapid prototyping, though Google says it can still produce outputs close to its larger models, with some tradeoffs in text, data accuracy, and character consistency.
- 6Hacker News·6/29/2026open_sourceHackerRank Open-Sourced Its ATS, Revealing LLM Scoring Variability
A Hacker News post discusses HackerRank’s open-sourced applicant tracking system and reports that the same resume can receive widely different scores across repeated runs. The author argues that LLM-based resume grading is highly non-deterministic in some categories, making hiring cutoffs unreliable.
- 7Hacker News·6/28/2026regulationGoogle reportedly restricts Meta’s access to Gemini models
Google has reportedly limited Meta’s access to its Gemini AI models, according to a CNBC report citing the Financial Times. The move highlights growing competitive tension between major AI companies over access to proprietary models and services.
- 7Google DeepMind·6/24/2026model_releaseGoogle DeepMind adds computer use to Gemini 3.5 Flash
Google DeepMind introduced computer use in Gemini 3.5 Flash, expanding the model’s ability to interact with software through a computer-like interface. The update is aimed at enabling more agentic workflows where the model can operate applications and complete tasks across screens.