Gemini felt dumb on May 19, 2026.

I was just exploring the new UI, poking through Gems and interactive visuals, when I ran into something totally bizarre. I suddenly got hit with a message saying my video generation limit was reached, even though I haven't made a single video! It’s so confusing and honestly a bit annoying to encounter such a glitchy, nonsensical error when I was just trying to check out the new features.

I can’t believe how much Gemini's image generation has been nerfed lately! It’s honestly insane; in just one week, the quality has plummeted so much that it feels like I'm using a model from a year ago. I tried to recreate some of my previous prompts, and the results were so poor that it left me completely shocked. It's incredibly frustrating to see such a massive drop in performance.

I'm so incredibly frustrated with Omni Flash. Every video I attempt to generate is met with a 'policy violation' error! I was just trying to swap a character into a clip of myself dancing, but the tool is being way too hyper-sensitive. Even when a prompt actually works, it fails 75% of the time if I try to reuse it. It’s basically impossible to use my credits!

I thought I finally had Gemini figured out. It was actually doing a great job breaking down my bloodwork patterns, but then—bam!—it deleted everything mid-sentence and hit me with a refusal. It’s incredibly frustrating because it clearly has the capability, but the guardrails just nuke the response. I've tried every hypothetical workaround, but the filter is just unyielding.

I am so incredibly annoyed with this. Google claims to have cutting-edge knowledge, but Flash 3.5 is practically a relic! It’s somehow 18 months behind on world events, which makes it feel like I'm stuck using some deprecated, useless model. It’s honestly ridiculous that a frontier model can't even keep up with reality. I expected so much more, but this is just a massive letdown.

I am absolutely floored by what Gemini Omni can do! I just threw a few screenshots at it, and it managed to one-shot the whole task perfectly. It felt almost surreal to see such high-level reasoning from such minimal context. I didn't have to prompt it repeatedly or fix errors; it just nailed it immediately. This level of precision is honestly mind-blowing and feels like magic.

I feel like I'm stuck in a time loop! It’s already May 2026, yet this AI is still stuck in January 2025. It is so frustrating to realize how outdated its knowledge is when I need real-time context. It feels like I’m talking to someone who hasn't checked the news in months. The tool's limitation is really holding me back; someone seriously needs to update that calendar!

Ever since that last Gemini update, the new UI has been driving me nuts. I've realized I can't even use the standard Android hamburger menu swipe to open the left sidebar anymore! I'm curious if anyone else is dealing with this. It feels like a total step backward in usability, and the way the app is behaving now is just incredibly frustrating and clunky to navigate.

I've been developing Llmisp, a tool that turns English specs into Clojure via a JSON AST. It’s been such an exciting project to work on! I’ve recently added GUI and SQLite support, and the results are honestly impressive. Seeing it generate over 650 lines of working code with a functional UI was a massive win for me. I'm really pushing to see how far we can take edge models like Gemma 4.

I used to love how Whisk handled accents; it was honestly amazing how it captured them. But now, using VEO 3.1 with Google Flow, I’m finding it incredibly frustrating. No matter how clearly I request a specific accent, the tool just completely ignores me and defaults to the same generic one every time. It’s such a disappointing step backward compared to my previous experience.

I'm feeling so defeated. I tried a video generation that completely flopped, and instead of getting nothing, it still snatched away 5% of my 5-hour limit! It's honestly devastating to be penalized just because the tool failed to deliver. It feels like dark times are ahead when every mistake costs me precious time—it's just so incredibly frustrating and feels totally unfair.