Three Years from GPT-3 to Gemini 3 by Ethan Mollick

Three years ago, we were impressed that a machine could write a poem about otters. Less than 1,000 days later, I am debating statistical methodology with an agent that built its own research environment. The era of the chatbot is turning into the era of the digital coworker. To be very clear, Gemini 3 isn’t perfect, and it still needs a manager who can guide and check it. But it suggests that “human in the loop” is evolving from “human who fixes AI mistakes” to “human who directs AI work.” And that may be the biggest change since the release of ChatGPT.

Google announced Gemini 3.0 which takes it closer to the state of the art with respect to other models. They claim it’s better than the rest. In this field, that’s a little subjective.

It has given me an interesting headache though. I was planning to take yearly subscription of Claude. I will test this out instead now.

Cloudflare explains Tuesday’s outage that temporarily took down ChatGPT by Richard Lawler

the query change caused its ClickHouse database to generate duplicates of information. As the configuration file rapidly grew to exceed preset memory limits, it took down “the core proxy system that handles traffic processing for our customers, for any traffic that depended on the bots module.”

My website is hosted on Cloudflare pages. It was down for a bit. As were a bunch of other websites - udemy, safari (o’Reilly) and so on.

It seems like a bad time for these cloud providers. First it was AWS, then Azure, then Azure had a DDoS attack and now this.

It truly seems like a matter of when and not if.

Microsoft Mitigated the Largest Cloud DDoS Ever Recorded, 15.7 Tbps - Slashdot by

On October 24, 2025, Azure DDoS Protection detected and mitigated a massive multi-vector attack peaking at 15.72 Tbps and 3.64 billion pps, the largest cloud DDoS ever recorded, aimed at a single Australian endpoint. Azure's global protection network filtered the traffic, keeping services online. The attack came from the Aisuru botnet, a Turbo Mirai-class IoT botnet using compromised home routers and cameras. The attack used massive UDP floods from more than 500,000 IPs hitting a single public address, with little spoofing and random source ports that made traceback easier.

I know you don’t want them to want AI, but… - Anil Dash by Anil Dash

I don’t know why today’s Firefox users, even if they’re the most rabid anti-AI zealots in the world, don’t say, “well, even if I hate AI, I want to make sure Firefox is good at protecting the privacy of AI users so I can recommend it to my friends and family who use AI”. I have to assume it’s because they’re in denial about the fact that their friends and family are using these platforms. (Judging by the tenor of their comments on the topic, I’d have to guess their friends don’t want to engage with them on the topic at all.)

AI is not going anywhere, so we better contribute to better ways of using the tool.