Archive for March 3, 2023

Friday, March 3, 2023

Brave Search Summarizer

Brave (via MacRumors):

The Summarizer provides concise and to-the-point answers at the top of Brave Search results pages, in response to the user’s input, solely based on Web search results. Unlike a purely generative AI model, which is prone to spout unsubstantiated assertions, we trained our large language models (LLMs) to process multiple sources of information present on the Web. This produces a more concise, accurate answer, expressed in coherent language.

In addition, the provenance of original sources of data is cited at all times via links. This maintains the rightful attribution of information, and helps users assess the trustworthiness of the sources, both of which are needed to mitigate the authority biases of large language models.

Previously:

OpenAI Is Today Unrecognizable

Chloe Xiang (via Hacker News)::

This blog post and OpenAI’s recent actions—all happening at the peak of the ChatGPT hype cycle—is a reminder of how much OpenAI’s tone and mission have changed from its founding, when it was exclusively a nonprofit. While the firm has always looked toward a future where AGI exists, it was founded on commitments including not seeking profits and even freely sharing code it develops, which today are nowhere to be seen.

OpenAI was founded in 2015 as a nonprofit research organization by Altman, Elon Musk, Peter Thiel, and LinkedIn cofounder Reid Hoffman, among other tech leaders. In its founding statement, the company declared its commitment to research “to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.” The blog stated that “since our research is free from financial obligations, we can better focus on a positive human impact,” and that all researchers would be encouraged to share “papers, blog posts, or code, and our patents (if any) will be shared with the world.”

[…]

By March 2019, OpenAI shed its non-profit status and set up a “capped profit” sector, in which the company could now receive investments and would provide investors with profit capped at 100 times their investment. The company’s decision was likely a result of its desire to compete with Big Tech rivals like Google and ended up receiving a $1 billion investment shortly after from Microsoft.

Previously:

Update (2023-03-16): James Vincent:

Yesterday, OpenAI announced GPT-4, its long-awaited next-generation AI language model. The system’s capabilities are still being assessed, but as researchers and experts pore over its accompanying materials, many have expressed disappointment at one particular feature: that despite the name of its parent company, GPT-4 is not an open AI model.

[…]

“I think we can call it shut on ‘Open’ AI: the 98 page paper introducing GPT-4 proudly declares that they’re disclosing nothing about the contents of their training set,” tweeted Ben Schmidt, VP of information design at Nomic AI, in a thread on the topic.

Previously:

Update (2023-03-24): Rohan Goswami (via Hacker News):

Altman didn’t take an equity stake in the company when it added the for-profit OpenAI LP entity in 2019, Semafor reported Friday.

Linode Price Increases

Shawn Michels (Hacker News):

The price of Shared and Dedicated compute plans will increase by 20%. Our shared Nanode plan remains unchanged at US$5 per month.

Additional IPv4 IP addresses will increase from US$1 to US$2 per month. IPv6 addresses remain free.

[…]

Akamai faces the same macroeconomic headwinds of growing costs for data center space, energy, and hardware as other providers. We resisted making any changes to our offerings for as long as we could.

csnover:

The message being sent to customers claims that they “remain more aggressively priced than our peer providers”, which is nonsense as far as I can tell—unless they now consider their “peer providers” to be the likes of AWS, Azure, and GCP?

nerdjon:

What is really weird about this, they are now more expensive than comparable systems on AWS Lightsail.

Marco Arment:

Deleted my earlier post about packet loss to Linode — that may be a regional Verizon thing closer to me, so I can’t blame them for it.

The price hike still has me looking around and making a contingency plan, though.

room34:

I’m not as concerned about the price increases as I am about Akamai just completely changing what Linode is, which I feel is inevitably underway. Very frustrating since I just moved to Linode last year after finally getting fed up with Digital Ocean’s laissez-faire attitude turning their network into a spam factory. Up to now I’ve been really happy with Linode.

Previously:

NSViewController.ViewLoading

NSViewController.ViewLoading:

Use this property wrapper on view controller properties that can be nil before the view controller’s view loads. Wrapping view controller properties this way eliminates crashes that can occur from implicitly defining properties as Optional, and then referencing them before the view controller finishes loading.

[…]

Use this property wrapper over implicitly unwrapped optionals for IBOutlets as well.

This is introduced in the Xcode 14.3 beta but requires macOS 13.3 (still in beta) for deployment. So it will be a long time before many developers can use this, but it seems like the sort of thing you could reimplement yourself.

There are also NSWindowController.WindowLoading and UIViewController.ViewLoading.

Previously:

Update (2023-04-22): Daniel Jalkut:

MagicLoading is a reference implementation of a @ViewLoading-style property wrapper.

[…]

The magic in Apple’s wrappers is an under-documented but widely used alternative mechanism for property wrappers: the ability to define getters and setters with knowledge of the containing object’s type, and access to the specific instance. The “subscript” variant of property wrappers is described in the original proposal, but has never been publicly endorsed, as far as I know, by Apple or the Swift team.

Usually I would shy away from undocumented features, but in this situation I believe it’s safe to use the undocumented subscript functionality because doing so only instructs the compiler to generate wrapper code that is compiled into the resulting binary. In other words: there are no private runtime dependencies that are being assumed by opting into this more sophisticated form of property wrapper.

So it shouldn’t count as private API for the App Store. I guess the main danger is that future versions of the compiler could remove support for this feature. Hopefully they will instead replace it with an official feature that’s more ergonomic.