pos

[Graham Stephan @GrahamStephan]

If you invested $1,000 with Warren Buffett in 1965, it would have grown to $4.3 million by 2010.

However, if Berkshire had been a hedge fund charging 2 & 20, that $4.3 million would have accrued $300k to the investor with a stunning $4m to the manager.

[pos @thisispos]

You might have heard about OpenAI's Q-Star Project, let's jump into it 👇

Introduction
In the rapidly evolving field of artificial intelligence (AI), OpenAI's Q-Star (Q*) project represents a significant milestone. Amidst a sea of speculation and rumors, it's essential to focus on the factual aspects of what Q-Star is and what it could potentially mean for the future of AI and cybersecurity.

Understanding Q-Star
Q-Star is an advanced AI model developed by OpenAI. Unlike its predecessors, Q-Star reportedly has capabilities that could revolutionize the field of AI. These include advancements in processing complex algorithms and potentially breaking modern encryption methods. However, it's crucial to approach these claims with a balanced perspective, understanding both the potentials and limitations of current AI technology.

The Implications for Encryption
One of the most discussed aspects of Q-Star is its alleged ability to break encryption. Encryption is the backbone of digital security, safeguarding everything from personal data to national security secrets. If Q-Star can indeed break encryption as claimed, this could have profound implications for cybersecurity. It raises questions about the readiness of our current systems to withstand such advancements and the ethical responsibilities of AI developers.

AI and Ethics: Navigating the Future
The development of Q-Star brings to the fore critical questions about ethics in AI. As AI models become more powerful, the need for responsible and transparent development becomes paramount. This includes considering the implications of AI on privacy, security, and societal structures.

Collaboration and Regulation
Q-Star's development also highlights the importance of collaboration between private entities like OpenAI and public institutions. The intersection of government projects and private research in AI development suggests a need for clear regulations and ethical guidelines to ensure beneficial outcomes for society.

Conclusion
OpenAI's Q-Star project, while surrounded by speculation, represents a significant step in AI research. As we venture further into this uncharted territory, it's vital to maintain a balanced and factual perspective, understanding both the opportunities and challenges posed by such technological advancements.

Reflective Questions
- How should the AI community balance innovation with ethical considerations in projects like Q-Star?
- What measures can be implemented to ensure that advancements in AI are used responsibly and for the greater good?
- How can collaboration between private and public sectors in AI research be optimized for societal benefit?

[Ben's Bites @bensbitesdaily]

Wow what a weekend... sheeesh!

Here’s what we have today in (Open)AI:

1/ OpenAI fired its CEO Sam Altman on Friday. Since then, combined seasons of Succession and Silicon Valley have happened.

2/ Sam Altman, Greg Brockman will join Microsoft with other top researchers from Open AI.

3/ Emmett Shear (co-founder of Twitch) as the Interim CEO of Open AI.

4/ Ilya Sutskever apologizes for his participation in the actions of the board a few minutes ago!

Our summary of the whole drama: https://t.co/XhqM2OhuUK

[pos @thisispos]

"When considering new capabilities such as AI, product managers should focus on how they can expand the user base, enhance users’ abilities, or eliminate tasks entirely." https://t.co/MtdTBtbO6a

[pos @thisispos]

Tech Live 2023: What’s Next for OpenAI?

https://t.co/rJVEPalpOw

[pos @thisispos]

The obscure court case that every big tech company is watching. https://t.co/itLlSk9nJd

[Alie Jules @aliejules]

Midjourney vs. Dalle-3 vs. Adobe Firefly

Same prompt for all.

10 spicy examples.

1.

[pos @thisispos]

We love stories about builders!

[Tibo @tibo_maker]

This guy makes $1M+ per year with 0 employees.

He started a side project which now has 500K+ users.

He spends only $700/year for maintenance & keeps ALL the profit for himself.

Ivan Kutskir, the world's most underrated indie-hacker 🧵

[Moritz Kremb @moritzkremb]

I spent 500+ hours in the last 6 months mastering ChatGPT.

The most effective way to improve your output is by using prompt frameworks.

Here are my top 5 prompt frameworks to help you level up your prompts:

[pos @thisispos]

Want an overview of LLMs in 2023? https://t.co/yaI7qRuEkp

[pos @thisispos]

Replit just made their AI coding assistant free for everyone. https://t.co/kvrpDT4SJI

[pos @thisispos]

Effective Accelerationism is the new Effective Altruism https://t.co/tlDVDEWkBv

[pos @thisispos]

The G7 will ask AI firms to agree to new safeguarding rules. https://t.co/r4Bb7pk1sK

[pos @thisispos]

@sama Nice!

[Alie Jules @aliejules]

Midjourney Office Hours - Aug 2, 2023

A quick recap.

v5.3
- still waiting on Discord to do the UI for the inpainting
- will include inpainting and a new style (aesthetic)
- release date: maybe about 2 weeks

v6
- had a candidate for v6 - It's better, but want it even better before the release
- more coherent, visually more diverse, better understanding
- working on making it more responsive to text
- instead of shipping out a more conservative v6, waiting to ship out the much better bigger v6
- release date: about a few weeks

Website
- working on the stand-alone website
- the biggest question at the moment is how much social stuff to put in and how much to change the front page
- a lot is getting better
- thinking about what the rollout of the new website will look like

Mobile App
- testing
- native iOS app coming
- non-native Android app coming
- website will also work on mobile!

Other
Updating their data centers

Image generation could still be faster
- thinking about how to do this
- faster computers but a limited amount
- possible try faster image generation sometime in the next month to make image generation a lot faster for pro and mega tier users (don't have enough computers to make it faster for all tiers)

Note.
Anyone (all tiers) can try Turbo mode right now, so you can try it. It's 4x faster.
You can add --turbo to the prompt or /turbo to switch to turbo mode.

Surveys
- check out and participate in the surveys in Midjourney discord; it's helpful information for them

- website will have surveys in the future, and when you participate, you'll get e.g a free GPU hour - right now, you can participate on Discord

Quote from David:
"Just make cool stuff."

Love that!!!

[pos @thisispos]

“We really didn’t have a choice at that point, delisting every asset other than bitcoin, which by the way is not what the law says, would have essentially meant the end of the crypto industry in the US,” he said. “It kind of made it an easy choice . . . let’s go to court and find out what the court says.”

https://t.co/45aZp2HcIB

[pos @thisispos]

Good thread!

[Misha @mishadavinci]

By 2030, tokenization of real-world assets will be a $16 TRILLION industry.

It will entirely transform real estate, private equity, stocks, bonds, and most forms of investing.

Here's what you need to know:

[pos @thisispos]

The origins of the use of [x] in UI design. ❎

https://t.co/uHgiSfKI6c

[Min Choi @minchoi]

AI just changed the game in filmmaking 📽️

Creators can now generate cinema quality short films, trailers, and teasers with 100% AI in just a few hours 🤯

10 favorite examples that will blow your mind:

[pos @thisispos]

Adobe’s new Generative Expand in Photoshop uses image generation AI to let users expand existing images. If there’s a bit of a pic missing, if it isn’t centred correctly, or if you just wanna make it bigger, the tool will use AI to help you do it. https://t.co/oQ7jLV7GN2