Weekend Warp

Where does one even start? I have refrained from writing the last two Weekend Warps, since it would have required to write a book each week. I do not think AI, or any other tech, for that matter has seen a December like the one this year, where every single day drips with meaty announcements.

So, let me restrict myself to meta points from the last couple of weeks:

谷歌 is back, and how! OpenAI went about creating its ’12 days’ (more below), but Google snuck in and stole the show. The big announcement was Gemini 2.0, of course, a vastly improved model, which caught up, and in some cases, upstaged the later versions of ChatGPT. What were more interesting were the products launched as adjuncts to Gemini 2 – Astra, which previewed an AI personal assistant; Mariner, (More here) which heralded Agentic AI frameworks; Gemini 2.0 Flash and Jules for developers (Read more here). ?It also previewed the powerful Deep Research, a personal AI research assistant, which will take Google Scholar to unimagined heights (see here). For good measure, it also released the next version of its breakout product NotebookLM, with additional bells and whistles and even a pro version (see here). In fact, in my opinion, Google’s zeitgeist started to change with this sleeper hit- NotebookLM like no other said that ‘Google is back’. Even Google-baiter, Perplexity ’s Aravind Srinivas had grudging words for praise for it (See here).

My AI&Beyond cofounder has done a nice job summarising Google’s stunning December releases in a post here.

I have also penned a provocation piece on ‘Why Google will own AI in 2025’ which will be published next week. Watch this space!

We are not finished with Google yet. An announcement as big, if not bigger, than the AI announcements were around Willow, its latest quantum computing breakthrough. Imagine if we had a super computer when the planet was created and we started doing a calculation so complex that the supercomputer would be still working away at it, knowing it would take 10 septillion years to crack it. Well, Google claims that Willow can do it in 5 minutes. Take that a with a fistful of salt, it is still worth writing home about! Quantum computing is a notoriously fickle technology requiring one to bend the laws of physics, and Google solve one huge part of it, by making error-corrected qubits get exponentially better as they get bigger. There is plenty of scepticism around this, but it gives the first real window into the possibility that Quantum Computing might be feasible in our lifetimes, hoping to solve intractable problems around climate change, logistics optimisation and others. (Read more here); and read an article I had in Mint many years back which demystified quantum computing. (Read here)

Ok, so let us now get to OpenAI. Its 12 days started with a bang, with the final release of Sora, the text-to-video generator (see here). It impressed, but was soon upstaged by Google’s Veo (see here), not yet generally released but looking much better in terms of video quality and other parameters than Sora. To its credit, OpenAI made an announcement every single day which included 1-800-CHATGPT, GPT Search, Apple Intelligence, Projects within ChatGPT, Canvas,etc. However, two announcements impressed. The first was amazing Advanced voice with video in ChatGPT (More here) which would help ChatGPT ‘see’, but Google launched something similar to crash this party too (see here). However, on the 12th day, OpenAI previewed ChatGPT o3, which, many say, took AI closer to AGI. O3 takes even more time to deliberate over questions—just a day after Google announced its first model of this type. OpenAI’s new model, called o3, replaces o1, which the company introduced in September. Like o1, the new model spends time ruminating over a problem in order to deliver better answers to questions that require step-by-step logical reasoning. (OpenAI chose to skip the “o2” moniker because it's already the name of a mobile carrier in the UK!) Santiago (@svpino on X) has a good 30 second take here.

Also, a great summary by Anuj Magazine here.

One last thing about Google: Riding high on the massive announcements, Google CEO Sundar Pichai took the momentum further by slashing 10% of its workforce. The start of AI and AI agents taking human jobs? 2025 will tell us more.

The space is certainly heating up...... at warp speed.