Latest / Neural Newscast / Google Renames NotebookLM and Debuts Gemini 3.5 Pro [Model Behavior]
Transcript
- [00:00] Nina Park: From Neural Newscast, this is Model Behavior, [00:03] Nina Park: AI-focused news and analysis on the models shaping our world. [00:11] Nina Park: Welcome to Model Behavior. I am Nina Park. [00:15] Nina Park: Today, we examine how AI systems are built and deployed in professional environments. [00:21] Nina Park: On this Friday, July 17, 2026, Google has initiated a major realignment of its product [00:27] Nina Park: suite. [00:28] Nina Park: This includes a rebranding of its most successful research tool and the release of a long-awaited [00:34] Nina Park: frontier model that had been delayed by technical hurdles. [00:37] Nina Park: I am Thatcher Collins. [00:39] Nina Park: It has been a busy morning for the teams at Google DeepMind and Labs. [00:43] Nina Park: We are looking at two major developments, the transition of Notebook LM into Gemini Notebook, and the arrival of Gemini 3.5 Pro. [00:52] Announcer: This rollout is significant because Google is playing catch-up on a release schedule that slipped by six weeks over the last month. [00:59] Announcer: Let us start with the product evolution. [01:01] Announcer: Google is renaming Notebook LM to Gemini Notebook. [01:05] Announcer: This tool started as Project Tailwind back at I.O. 2023 and has grown to 30 million users and 600,000 organizations. [01:15] Announcer: According to Google, while it remains a standalone product, the integration into the broader Gemini ecosystem is becoming explicit, moving away from its experimental roots into a core productivity tool. [01:27] Thatcher Collins: The technical shift is more than just a name change, Nina. Google is introducing a secure cloud computer for every notebook. [01:35] Thatcher Collins: This allows Gemini Notebook to natively write and execute code. [01:40] Thatcher Collins: For researchers, this means you can perform complex data analysis that is grounded specifically [01:45] Thatcher Collins: in the sources you have uploaded, rather than relying on the general knowledge base of the [01:50] Thatcher Collins: model. [01:51] Announcer: That is a major shift for a tool once seen as a simple summary engine. [01:56] Announcer: It is moving into the realm of active computation. [01:59] Announcer: However, these features are not universal yet. [02:02] Announcer: The integrated code execution is available today for AI Ultra and Workspace business users, [02:09] Announcer: while the rollout to standard Pro users on the web is scheduled for the coming weeks. [02:13] Thatcher Collins: It is a clear monetization play. [02:16] Thatcher Collins: But the bigger technical story today is Gemini 3.5 Pro. [02:20] Thatcher Collins: It officially launched this morning, July 17th. [02:23] Thatcher Collins: This is the model promised back in June, and as reported, it arrives roughly six weeks late. [02:29] Thatcher Collins: It is landing in a market where GPT-5.6 on and Claude Fable 5 have already been setting the pace for several weeks. [02:37] Announcer: The standout specification for 3.5 Pro is the context window. [02:42] Announcer: We are looking at 2 million tokens. [02:45] Announcer: To put that in perspective, Thatcher, that is 10 times the capacity of Claude Fable 5. [02:51] Announcer: Google is positioning this as the premier tool for professionals who need to process [02:55] Announcer: entire code bases or massive research libraries within a single prompt. [03:01] Thatcher Collins: The context window is impressive, but we have to look at the reasoning layer. [03:05] Thatcher Collins: Google calls DeepThink. [03:07] Thatcher Collins: This is their answer to the complex, multi-step problem-solving seen in recent competitor [03:12] Thatcher Collins: releases. [03:13] Thatcher Collins: But here is the friction. [03:14] Thatcher Collins: DeepThink is gated behind the Ultra subscription tier, which costs $250 per month. [03:21] Thatcher Collins: That is a high barrier to entry compared to other models. [03:24] Announcer: It's also worth noting why this took so long. [03:27] Announcer: Reports indicate DeepMind had to scrap the original Gemini 2.5 Pro architecture entirely. [03:33] Announcer: They encountered structural failures in recursive tool calling and SVG scene generation. [03:39] Announcer: They basically had to re-build the foundation mid-cycle, [03:43] Announcer: because the previous iteration was not meeting the frontier performance standards set by their peers earlier this summer. [03:49] Thatcher Collins: That explains the delay, but it puts Google in a tough spot. [03:54] Thatcher Collins: They are the fourth major lab to ship a frontier model this season. [03:58] Thatcher Collins: While the 2 million token window is a unique selling point, the market has already had time to stabilize around other ecosystems. [04:06] Thatcher Collins: Google is asking users to switch to a more expensive tier for a model without a track record. [04:12] Announcer: It is a high-stakes move to see if sheer context size can outweigh a late arrival and premium pricing. [04:20] Announcer: For professionals who need to analyze massive data sets, that context window might be the deciding factor. [04:26] Announcer: But for general reasoning, the competition remains fierce. [04:30] Announcer: We will see if this realignment helps Google regain its footing in the frontier model race. [04:36] Announcer: Thank you for listening to Model Behavior, a Neural Newscast editorial segment. [04:41] Announcer: You can find more analysis at mb.neuralnewscast.com. [04:46] Announcer: Neural Newscast is AI-assisted, human-reviewed. [04:49] Announcer: View our AI transparency policy at neuralnewscast.com. [04:54] Nina Park: This has been Model Behavior on Neural Newscast. [04:57] Nina Park: Examining the systems behind the story.