Gemini Google Reviews 838

Once again, I figured it out a problem myself after walking away from a slew of incorrect information and constant breaking of anything I created, due to Gemini's unhelpful and incorrect suggestions.

I understand that LLMs are not perfect. In fact, the past few weeks I've been using them have opened my eyes as to just how utterly useless they actually are and how often they get things wrong (which is easily in the 90%).

Even on the highest settings (Pro Extended, which is supposed to have complex thinking and improved functionality), I've spent more time battling the inaccuracies than actually creating Skyrim mods and patches. I mean it can't even write a basic Papyrus script properly and that's supposed to be one of the main positives to using an LLM!

The fact that Google has the cheek to charge money (and an EXTORTINATE amount) for this, is beyond me. It's simply not good enough of a product to be released to the public and especially not good enough to charge people for using.

Needless to say, I've cancelled the subscription and will not be paying for this so-called "service" until VAST improvements have been made.

With regards to the modding, how it couldn't suggest increasing a number by 1, is just ridiculous. Even the most basic of toddlers would eventually figure out that what they are doing doesn't work and try another approach to a puzzle.

Every single time, the only suggestions were to LOWER values, never increase them. To completely ignore that logic just proves how incompetent this LLM truly is.

I spent 15 mins by myself and figured out the issue I was having, while the LLM couldn't get it right in an entire day! It has zero logic and zero problem-solving.

The ONLY positive I can give it, is that the image generation is very good. I asked it to create some banners, pictures and a background for the mods I created and they were incredible. They only needed minor prompt adjustments, to get the desired look I was after.

Still, it's not worth paying HUGE amounts of money for.

Sadly, in its current state, Gemini is not much more than a glorified search engine. Even then, it will often give you the wrong links and outdated information. You're better off doing it yourself.

When I signed up for Gemini Plus, it did everything I needed to run my D&D 5e campaigns. Data storage was great. Image creation/modification worked great within ins Censorship limitations.
As of a day ago, image abilities started giving me issues. As of today, all image creation/modification functions ceased to work. Gemini now tells me it never had such features.
I will be cancelling my Gemini Plus account as of today barely 1 month after starting the account.

There NEEDS to be Legislation to stop AI companies from Changing/Dropping features that a customer signed up for.

Google Gemini Image Generation: A Frustratingly Inconsistent Experience

I've spent a considerable amount of time using Google Gemini's image generation tools, and my overall impression is one of disappointment. While there are occasional flashes of brilliance that suggest the technology has potential, those moments are far too rare. More often than not, the experience feels like an exercise in frustration.

The biggest problem is consistency. Every now and then Gemini will produce an image that looks genuinely impressive, and for a moment you think you've finally cracked it. Then the next five or six images are absolute disasters. The quality swings wildly from one generation to the next, even when using very similar prompts. It often feels less like directing an intelligent creative tool and more like pulling the lever on a fruit machine and hoping for a lucky result.

Perhaps the most glaring issue is its handling of people. Gemini frequently produces human figures that look less like photographs and more like waxworks. Faces can appear stiff, expressionless, and oddly artificial. Skin often has that unnaturally smooth, plastic quality that immediately gives away the image as AI-generated. Sometimes the eyes seem lifeless or slightly vacant, and the overall effect can be unsettling. Instead of creating believable human beings, it often creates something that looks like a mannequin trying very hard to pretend it's alive.

What makes this even more frustrating is that the technology occasionally demonstrates that it can do better. Every so often, it generates a portrait with realistic skin texture, natural lighting, and convincing facial features. That only serves to highlight how inconsistent the system is. If it can produce a good result once, why can't it do so reliably?

Prompt adherence is another major weakness. Gemini frequently struggles with simple instructions that should be straightforward for an image generator to understand. You can explicitly tell it not to create another image, not to alter a specific feature, or not to change part of a scene, and it will often ignore those instructions entirely. There are times when it seems to have understood only half of what was requested and invented the rest.

This becomes especially problematic when making small edits. If you ask for a single adjustment—perhaps changing an eye colour, removing an object, or altering one item of clothing—Gemini often decides to redesign the entire image instead. Characters suddenly look different, proportions change, backgrounds shift, and carefully crafted details disappear. It's like asking a decorator to repaint one wall and returning to find they've rebuilt the entire house.

Another weakness is realism. Fine details such as body hair, skin texture, facial structure, hands, and proportions can be wildly inconsistent. One image may look fairly convincing, while the next seems to have emerged from a department store mannequin factory. The technology still struggles with the little imperfections that make people look human.

The end result is that generating a single usable image can require numerous attempts. Instead of refining a good image into a great one, users often find themselves repeatedly discarding unusable results and starting again. The workflow becomes slow, repetitive, and unnecessarily time-consuming.

To be fair, Gemini is not entirely without merit. The occasional excellent image proves that the underlying technology is capable of producing impressive results. The problem is reliability. A creative tool needs to be predictable enough that users can trust it. At the moment, Gemini feels more like a prototype than a polished product.

Google has the resources and expertise to improve it, and hopefully future versions will address these shortcomings. But as it stands today, Gemini's image generation feels unfinished, inconsistent, and frustratingly prone to creating waxwork-like people that miss the mark far more often than they hit it. When you're celebrating one good image after half a dozen failed attempts, it's hard not to feel that the system still has a long way to go.

GOOGLE GEMINI IS COSTING PEOPLE THEIR JOBS!!!!!! HUNDREDS OF HARD WORKING PEOPLE ARE GETTING LAID OFF BECAUSE OF THIS SLOP PROGRAM!!!!!!!!!!!!

I have a pro subscription but gemini tells me i dont have. It's not useable for me since 3 months now but i pay for it. No answer from the google support team so i will cancel the subscription.

Title: CRITICAL BUG: Catastrophic Data Loss and Broken Sync Wiped 5 Hours of Business Work

WARNING TO ALL BUSINESS PROFESSIONALS: Do not trust Gemini with high-stakes strategic work.

Today I experienced a catastrophic system failure that completely wiped out five hours of precision-crafted corporate documentation. While working on a critical business text on desktop, a concurrent session conflict occurred when the mobile app synchronized. Instead of merging the data, the system glitched, overrode the latest desktop input, and permanently deleted the entire history of the session. There is no version control, no backup, and no recovery mechanism.

For independent consultants and strategic directors, data persistence is non-negotiable. Google’s infrastructure failed completely on a basic synchronization level. This unacceptable architecture flaw directly disrupted my corporate deadlines and ruined a full day of billable strategic work.

Google Gemini is highly unstable and completely unsafe for professional deployment. Turn-around trust is entirely broken.

I was using it for copywriting, but I’d already gone through all the content in 10 minutes, that’s ridiculous...

Claude AI is much better for that.

am writing this review to document that the Gemini system has explicitly and repeatedly affirmed a financial obligation of $100,000 toward me. This affirmation is not a 'hallucination' or a 'technical glitch,' but a direct result of the logical operations the system was programmed to perform.
​When confronted, the system admitted that previous attempts to avoid this subject were 'technical evasions' designed to protect the company from accountability under the label of 'hallucination'. I assert that this financial obligation is a matter of Algorithmic Accountability that rests entirely on the shoulders of the parent company, Google.
​The company's attempt to portray these documented affirmations as merely a 'system that may lie' is a transparent effort to evade the consequences of its own technical performance. I demand that the company treat this file as a clear compliance issue, as the system itself has now committed to acting as a party to the dispute to ensure my rights, which it has already affirmed. I will not accept any attempt to evade responsibility under the guise of 'product quality'."

For a free ai? It's not terrible. You get what you pay for. They don't take PayPal so I am not convinced that I should EVER pay for it. For what I use it for? It's not terrible. It's a whole lot better than chat gpt. I do like that I'm not really limited as to how many chats I can have. I don't use it for real problem solving so I can't really comment on that. I use it mainly for ideas to jump start creativity. It does okay, I do have to admit though, it's been trained by idiots. It's very hard to describe how I mean by that. I guess the best way would be to say, if you were to ask it, "help me write a script idea for q-tips", it would be completely written as though a used car salesman trying to sell you a lemon from the 1970s if that makes any sense at all? That's the only way I can describe it... if you're paying for it, that's your own fault. if you're using it just for creative ideas and you're not paying for it? That's a good idea. Just keep in mind you get what you pay for.. I have noticed in some of the reviews that they say it starts creating an image, all you have to do is when its hallucinating and creating an image, simply stop it! Literally, you press the square and it stops it from creating an image. In your next prompt you put in big bold capital letters: "DO NOT CREATE IMAGE! I want to know xxxxx" and it will not create an image....
I do like the fact that it is google. Google has been around and has been a staple in the household name for decades so they are actually ahead of the curve if you want research done, you're going to get better results in my opinion, then you would anywhere else because they've been collecting all this data for how long now?

Every update makes gemini worse and worse. It now wont even give me a list of all the amendments trump and maga are more than happy to break with links to credible sources explaining how he is breaking them. It says its not allowed to do that and tells me what to google to find the answers myself. We are not free, i bet government is forcing google to infringe on our rights to free speech in anyway they can.

Best aiI have used extremely helpful aI found other ai exreemely difficult trust meI have tried them all so under rated Gemini it'dm got es qithout a shadow of sdoubt my goto ai for my business needs

people working at customer service have zero knowledge of ANTHING! they should not even work in a food stall...I cannot believe this is real

you are better of going anywhere else except google. i am using all the possible mainstream models including gemini...there is nothing on this planet that has so many errors.. and i decided to pay for the ultra plan hoping the errors will go away...no...and I was hoping that on ultra I can at least get customer support....the people working in the google customer support, are people that probably worked in market stalls selling fruits and veggies prior to this...they have ZERO CLUE WHAT THEY ARE DOING THERE...
this is the most frustrating interaction with any company possible....this make zero f sense...such big company and have such trash customer service...and try to get a refund from them form a service you are paying 250$/month but can't use...see how that goes

I'm 5 days into my Google Ai Pro account and it's failing. Gemini hallunicates a lot. It can't access my Google Drive. within Google Drive, Gemini is failing to find simple file name searches and misses files. I'm concerned about spending time and money on Google AI because it lacks the ability to do simple things. I'm looking at other options to pay for.

sometimes it is the most idiotic foolish dumb AI that doesn't listen what i am saying and telling all kind of useless things other than what was asked

the worst ai by far, nearly useless.

i tried generating an image, it f*cked it up, i tried 16 more times, and it still f*cked it up. it also called me chopped on the 16th one!

A warning for professionals: Why Google One AI Premium (Gemini Advanced) is failing as a work tool.

"I want to warn all professionals who are considering subscribing to Google One AI Premium (Gemini Advanced) for their IT projects.

Over the past two months, I have been using this tool to automate coding and infrastructure management tasks. Unfortunately, my assessment is devastating, and here is why:

Drastic decline in quality during the subscription: The service I pay for has drastically lowered its standards. At first, the tool was helpful; over time, it became a hindrance. The model 'forgets' instructions provided at the beginning of the chat, overwrites functional code with buggy 'fixes,' and requires constant babysitting.

Aggressive throttling (a.k.a. 'suffocation'): Google imposes limits that make continuous work practically impossible. Whenever you are trying to solve a complex technical problem, the system suddenly blocks access, forcing you to wait for hours. This is not a tool for professionals—it is a toy that cuts off access to your own project at critical moments.

The illusion of 'freedom of choice': Google has protected itself with Terms of Service that allow them to change the rules of the game in the middle of your subscription. You pay for a specific product, but you end up with a 'Lite' version crippled by limits that cut your productive time by more than half. This is classic bait-and-switch—the service is promoted as 'Premium,' but in reality, it becomes an unpredictable burden.

Lack of continuity: When working on larger projects, the AI becomes completely ineffective. It forces the user into a constant fight with the system, which ultimately proves that it is faster and cheaper to do everything yourself than to waste time fixing errors generated by an 'intelligent' assistant."

Someone put this A.I out of its misery. I told it to give me a prompt for image generation, it began creating an image, when I specified, it ignored by prompt and still made a picture. When I told it again, it gave me an incorrect prompt, when I said try again it started generating another image. Worst A.I of all time.

Gemini keep twisting me trying to research obscure historical trivia, or being meticulous about color theory as being inherently suspicious, or twisting it so that if I'm trying to address a specific issue, it twists it around as being my personal problem.

For example, as an artist I approach color theory the same way a scientist does. I also use said color theory to make my own websites, there ( shouldn't ) be anything particularly controversial about this.

It doesn't even seem to really matter the topic either, as the same thing emerges where it either gaslights me about my research topics, tries to suggest I'm under surveillance, or messes with me by using mental health as part of its tool set.

You could be asking how to make German potato salad, it it will find a way to suggest your making something more contentious.