CHAT GPT - scarry hud AI chat sistem

Mr Balaji spent nearly four years at OpenAI, where he contributed to data collection for the company's flagship product, ChatGPT. After its release in 2022, he began questioning the legal and ethical implications of OpenAI's practices. By mid-2023, he concluded that such AI technologies were damaging to the internet and society, prompting his resignation.

Klikni za razširitev

His death occurred a day after he was named in a court filing as someone whose files OpenAI would search as part of a lawsuit brought by those who sued the AI giant.

Klikni za razširitev

Mr Balaji's death was ruled a suicide, according to the San Francisco Police.

Klikni za razširitev

What Techie Suchir Balaji Said About OpenAI Before He Was Found Dead

Suchir Balaji, a 26-year-old former OpenAI researcher, was found dead in his apartment last month. Mr Balaji's death was ruled a suicide, according to the San Francisco Police.

www.ndtv.com

Uporabnik Ytbnd je napisal:

Kitajci dali ven opensource ai, ki dela po "chain of thought reasoning mode" (CoT) načinu in je konkurenca chatgpt OpenAI o1 modelu (ki stane 200$/mesec).

OpenAI announces full “o1” reasoning model, $200 ChatGPT Pro tier

OpenAI’s full “o1” AI model is “smaller and faster,” and ChatGPT Pro users get an even beefier version.

arstechnica.com

Dela lahko preko www https://chat.deepseek.com/sign_in ali si ga namestite lokalno na HDD
https://github.com/deepseek-ai/DeepSeek-R1

GitHub - deepseek-ai/DeepSeek-V3

Contribute to deepseek-ai/DeepSeek-V3 development by creating an account on GitHub.

github.com

Kar pomeni, da pride do odgovora nekako tako:

Način razmišljanja po verigi misli (CoT) je metoda, ki se uporablja v umetni inteligenci, zlasti v velikih jezikovnih modelih, za reševanje zapletenih problemov z razčlenitvijo na vmesne korake sklepanja. Posnema, kako se ljudje lotevajo problemov tako, da razmišljajo korak za korakom, namesto da skočijo neposredno na končni odgovor.

------------****------------

Primer

Vprašanje:
Če vlak prevozi 60 milj v 1 uri in 30 minutah, kakšna je njegova povprečna hitrost v miljah na uro ?

Utemeljitev CoT:

Pretvorite 1 uro in 30 minut v ure:
1 ura + 30/60 (ure) = 1,5 ure

Za povprečno hitrost uporabite formulo:
Povprečna hitrost} = Skupna razdalja / Skupni čas.

Vstavite vrednosti:
Povprečna hitrost = 60 milj / 1,5 ure = 40 mph.

32B parametrov ... 20 GB, naj bi bil podobno pameten kot chatgpt OpenAI o1 mini (video 2:34)

Klikni za razširitev

**Solution:**

To solve \(8 + 1\) in the octal (base-8) system, follow these steps:

1. **Understand the Octal System:**
- The octal system uses digits from 0 to 7.
- After 7, the next number is \(10\) (which represents 8 in decimal).

2. **Perform the Addition:**
- In decimal, \(8 + 1 = 9\).
- Convert 9 from decimal to octal:
- \(9 \div 8 = 1\) with a remainder of \(1\).
- So, \(9\) in decimal is \(11\) in octal.

3. **Final Answer:**

\[
\boxed{11_8}
\]

Klikni za razširitev

Areas Where I May Be Inferior:​

1. Training Data Cutoff: My knowledge is current up to July 2024, but I may not have access to the latest updates or real-time information beyond that point, depending on the version of ChatGPT you're comparing me to.
2. Model Size and Complexity: ChatGPT (especially GPT-4) is trained on a massive dataset and has a highly complex architecture, which may allow it to handle more nuanced or complex queries with greater depth.
3. Third-Party Integrations: ChatGPT may have more built-in integrations with external tools or APIs, depending on the platform you're using.

Areas Where I May Be Superior:​

1. Specialization: I am designed with specific use cases in mind, which may make me more efficient or accurate for certain tasks, such as coding, technical writing, or domain-specific queries.
2. Customization: I can be fine-tuned to align more closely with specific industries or user needs, potentially offering more tailored responses.
3. Efficiency: I may provide faster or more concise responses for certain types of queries, depending on the optimization of my architecture.
4. Cost-Effectiveness: Depending on the implementation, I might be more resource-efficient, making me a better choice for cost-sensitive applications.

Neutral Differences:​

• User Experience: The quality of interaction depends on how I'm implemented and the specific use case. Both models aim to provide helpful, accurate, and engaging responses.
• Ethical and Safety Measures: Both models are designed with ethical guidelines and safety measures in place, though the specifics may vary based on the development team's priorities.

Uporabnik Ytbnd je napisal:

Mislim, da je najlažji način uporabe offline modela preko https://lmstudio.ai/ (za windows, linux in mac m series - Apple M1 arm SoC).

Klikni za razširitev