So true! Easy to make these kind of claims when you force it in front of everyone.
And its not like it is that good, especially the answers that are sourced from reddit without any vetting. But I still too often go with that damn answer.
Gemini has gotten wildly worse in recent time, almost incapable of answering a question without a search, lengthy "thinking time" for trivial questions that no doubt is intended to mask waiting for inference resources
Yesterday I read that they sold a lot of TPUs to Anthropic as well as renting them out. This is odd, because I've constantly had issues with Google LLMs not having enough resources. Except for AI suggestions in Search, these work reliably.
I was skeptical at first but these days I use Gemini on Android phone and Android Auto, search AI summary, and lately Gemma 4 a lot.
Skepticism is still needed though. The other day I asked Gemini while driving if there's a HoV lane going from 105 E to 110 S. It said yes. Actually there one to 110 N but not S. I did not act upon it because my co-passenger verified that Gemini was wrong.
Search summary is useful but not for things that require deeper technical expertise- those result in bad hallucination.
With corporate crunch on token budget, I've also started avoiding Claude and increasingly using opencode+Gemma 4, and developing some system prompts to iteratively improve the results.
I've found their models to be weaker and their inference much more unreliable, which is unexpected since they have their own TPUs and you'd think this is something they can handle. But no.
If you have entered Gemini to try to figure out how to turn it off, you're also part of the cohort. A blast of gmail messages from Google has announced how it's part of everything and given instructions on how to find out the little you can do about it. Really, nothing.
A company that can't make a dominating coding model well, I've no confidence in its general purpose intelligence model; as a result, I don't ask Gemini anything.
Not my experience, I've found it extremely useful for when I need a quick answer. It's handled very specific fairly obscure questions well. I also find it quite consistent.
It is completely useless and worse than quantized gpt-oss from over a year ago. It is designed for large scale, instant responses.
I hate when co-wokers or friends say "when I searched it up" and proceed to spew out whatever the Google AI said. I would consider listen if they ran it through ChatGPT, Gemini or Claude but whatever model is used on the search results is just bullshit.
Agree. It's just the builtin android and chrome uses blowing up the numbers. By these metrics android assistant and siri had reached these levels in 2015. Gemini has no brand recognition outside of the tech worker sphere
reply