will_a113

joined 2 years ago
[–] [email protected] 3 points 1 week ago (1 children)

Not that we have any real info about who collects/uses what when you use the API

[–] [email protected] 4 points 1 week ago

Nobody knows! There's no specific disclosure that I'm aware of (in the US at least), and even if there was I wouldn't trust any of these guys to tell the truth about it anyway.

As always, don't do anything on the Internet that you wouldn't want the rest of the world to find out about :)

[–] [email protected] 4 points 1 week ago (3 children)

They're talking about what is being recorded while the user is using the tools (your prompts, RAG data, etc.)

[–] [email protected] 1 points 1 week ago

If money counts as a freedom unit then yes, probably (maybe)

[–] [email protected] 3 points 1 week ago

Anthropic and OpenAPI both have options that let you use their API without training the system on your data (not sure if the others do as well), so if t3chat is simply using the API it may be that they themselves are collecting your inputs (or not, you'd have to check the TOS), but maybe their backend model providers are not. Or, who knows, they could all be lying too.

[–] [email protected] 29 points 1 week ago (3 children)

And I can't possibly imagine that Grok actually collects less than ChatGPT.

 

With, I think, a massive grain of salt since this info is unverified and direct from the manufacturer...

Huawei’s official presentation claims their Cloudmatrix 385 supercomputer delivers 300 PFLOPS of computing power, 269 TB/s of network bandwidth, and 1,229 TB/s of total memory bandwidth. It also achieves 55 percent model fitting utilization (MFU) during training workloads and offers 2.8 Tbps of inter-card bandwidth, heavily emphasizing its strength in networking.

| Spec            | NVL72 (Nvidia) | CloudMatrix 384 (Huawei) | Better? (%) |
|-----------------|----------------|--------------------------|------------|
| Total compute   | 180 Pflops     | 300 Pflops               | 67%        |
| Total network bw| 130 TB/s       | 269 TB/s                 | 107%       |
| Total mem bw    | 576 TB/s       | 1,229 TB/s               | 113%       |
 

A chart titled "What Kind of Data Do AI Chatbots Collect?" lists and compares seven AI chatbots—Gemini, Claude, CoPilot, Deepseek, ChatGPT, Perplexity, and Grok—based on the types and number of data points they collect as of February 2025. The categories of data include: Contact Info, Location, Contacts, User Content, History, Identifiers, Diagnostics, Usage Data, Purchases, Other Data.

  • Gemini: Collects all 10 data types; highest total at 22 data points
  • Claude: Collects 7 types; 13 data points
  • CoPilot: Collects 7 types; 12 data points
  • Deepseek: Collects 6 types; 11 data points
  • ChatGPT: Collects 6 types; 10 data points
  • Perplexity: Collects 6 types; 10 data points
  • Grok: Collects 4 types; 7 data points
[–] [email protected] 5 points 2 weeks ago (1 children)

Ok so you’d literally be making a regular Lenny post to some particular community on some particular instance in that case, right?

[–] [email protected] 4 points 2 weeks ago (6 children)

I’m a little lost. You mention hosting content on any instance, or on GitHub. How does that work? And if your content is elsewhere what is Lemmy doing? Authx?

[–] [email protected] 3 points 3 weeks ago

Some of the instances have a community bent to them. Slrpnk and beehaw come to mind most immediately, but there are lots of others that are focused on gathering a specific niche of humanity and getting them to interact with one another.

[–] [email protected] 13 points 3 weeks ago (1 children)

Back in the old days (ie 2020) we called this “writing out requirements” and it was generally the way that devs knew what they were supposed to do.

[–] [email protected] 5 points 3 weeks ago (4 children)

I think Maid Marion from Disney’s Robin Hood basically invented furries.

[–] [email protected] 3 points 1 month ago (1 children)

Voyager is mobile-first, but can be run as a webapp frontend that you can host locally (out of a repo or docker container), or set up for web access. If it's just keyword filtering you're after you can also use ublock origin rules to do that (in fact there was a lemmy thread today talking about it somewhere) You can set rules like

lemmy.ml##.post-listing:has(.post-title:has-text(/trump|elon|musk|biden|kamala|rfk/i))

and the content will simply not render.

19
submitted 2 months ago* (last edited 2 months ago) by [email protected] to c/[email protected]
 

Crowds and water have more in common than you'd think - they both flow like a fluid, with predictable patterns that can turn perilous if not properly managed. Looks like the physics of human herds is no bull, as researchers have uncovered the fluid dynamics behind dangerous crowd crushes.

 

Using Reddit's popular ChangeMyView community as a source of baseline data, OpenAI had previously found that 2022's ChatGPT-3.5 was significantly less persuasive than random humans, ranking in just the 38th percentile on this measure. But that performance jumped to the 77th percentile with September's release of the o1-mini reasoning model and up to percentiles in the high 80s for the full-fledged o1 model.

So are you smarter than a Redditor?

 

When even Cory Doctrow starts to sound like an optimist I have to give myself a reality check as it usually means I'm heading off the deep end. But in this case it just rubs me the wrong way that he talks about Mastodon and Bluesky in the same breath -- one is not like the other.

 

Originality.AI looked at 8,885 long Facebook posts made over the past six years.

Key Findings

  • 41.18% of current Facebook long-form posts are Likely AI, as of November 2024.
  • Between 2023 and November 2024, the average percentage of monthly AI posts on Facebook was 24.05%.
  • This reflects a 4.3x increase in monthly AI Facebook content since the launch of ChatGPT. In comparison, the monthly average was 5.34% from 2018 to 2022.
 

And I just assumed they called Rainbolt

 

Hackers have reportedly breached Gravy Analytics, a parent company of Venntel that sells smartphone location data to the U.S. government. The hackers claim to have stolen considerable data, including customer lists, industry information and individuals' location data. They are threatening to make the data public.

 

To get around the problem of damaging 10,000 neurons just to connect with 1,000, Biohybrid is experimenting with an approach that makes donor neurons a part of the implant itself - potentially allowing for dramatically better connection scaling.

 

Hahahahahahahahahahahahaha

 

In this niche case the Vision Pro seems like it has some compelling benefits.

view more: next ›