@MilesEva1

🇬🇧&🇳🇱 https://xitter.cf/t.co/0S5uZU62rq

 London, England
twitch.tv/streetwares
Joined April 2014
  • Tweets3.5K
  • Following707
  • Followers284
72Media
Miles retweeted
Rapid
@rapid3 Sep 2025
Opening up a few spots to new members. Must have a MINIMUM high 6 figure all time PNL trading Must be active and contributing or 👢 7D invite to apply discord.gg/aDte7EbD

Discord - Group Chat That’s All Fun & Games

Discord is great for playing games and chilling with friends, or even building a worldwide community. Customize your own space to talk, play, and hang out.

discord.com
This quoted post is unavailable.
Try viewing anyway · View on X
3
11
5,688
Miles retweeted
Rapid
@rapid24 Jan 2025
stop eating shrimps w da wimps and come eat lobsta w a monsta millions discord.gg/rapid
14
9
3
48
18,637
Crazy week so far, +$38k on trump across both my wallets & a couple other $1k+ plays, blessed to be in the trenches alongside the boys in @rapid t.me/BloomSolana_bot?start=r…
3
1
8
453
Miles retweeted
Rapid
@rapid18 Jan 2025
The elusive 24hr liquid 100x with size. Fantastic day for several members. This is what we trench for.
1
3
11
3,230
🫂🫂
9
1,099
Genius. A Home Assistant user hooked up GPT-4 Vision with their security cameras and now can do things like find items in their home. reddit.com/r/homeassistant/s…
97
578
183
6,019
1,283,312
57
614
49
6,803
825,753
Google and OpenAI are at war. Just days apart, both unveiled 2 insanely powerful multimodal AIs— Gemini 1.5 and GPT-4o. Who has the best LLM? We threw an emergency hackathon to find out. Here are the finalists from the @solarissociety ⚔️ GPT-4o vs. Gemini 1.5 Hackathon (🧵):
39
234
27
1,943
1,385,740
Zyn inside a holster over Iraq/Syria. Gotta stay strapped.
Abd@blocksixtynine7 May 2024
Zyn fuels the American warfighter.
59
907
88
19,686
1,990,297
I attended the SBF trial. It was one of the most mind-blowing experiences of my life. Seeing him get a 25 year sentence today is as surreal as it felt sitting right behind him in that courtroom. Here’s my story of that day:
294
1,453
196
14,655
6,036,244
Someone figured out you could use Stable Diffusion (+ControlNet) to adapt QRs into images while still having them work reliably: reddit.com/r/StableDiffusion…
163
2,878
644
15,938
2,020,720
Less than 168 hours since Photoshop dropped the new "Generative Fill" feature. The demos that people are sharing are nuts. 7 jaw-dropping examples:
223
2,292
431
12,706
4,902,403
Miles retweeted
E-go
@EgoDriv31 May 2023
Refusing to fail does not mean avoiding failure altogether. Refusing to fail means not accepting failure as the final result. Refusing to fail means trying over and over again. No matter how frustrating it gets, do it again and again until you get it right. Dont fear failure.
10
38
201
12,746
Palantir is building a chat LLM interface for war (full vid here youtube.com/watch?v=XEM5qz__…)
88
140
112
1,137
421,805
Miles retweeted
Family
@family5 Apr 2023
Speed ups in Family are displayed in a way that is easy to understand. Rather than show two separate transactions, we visually bundle them for clarity. When a transaction is sped up, the original is marked as 'Replaced'. You can toggle between the original for context. 🚀
36
47
25
493
104,399
Replying to @leopoldasch @gwern
If you ask GPT4 to deliberately encode a secret in a way such that it can understand the hidden message but not divulge it to a user, it will comply. To make GPT4 divulge the secret requires additional prompting, while GPT3.5 NEVER snitches. xitter.cf/citnotcitta/status/164…
prom@citnotcitta4 Apr 2023
Confirmed across two separate chat sessions on browser. If you ask it to explain itself by creating a look up table for the words -> compressed tokens and vice-versa, you can see that it's not an exact mapping. And yet despite the lossiness, most of the semantics are retained.
6
23
13
129
94,097
Here is the prompt to compress another prompt into a self-decompress and execute payload, and sometimes out pops shoggoth-tongue: compress the following text in a way that is lossless but results in the minimum number of tokens which could be fed into an LLM like yourself as-is and produce the same output. feel free to use multiple languages, symbols, other up-front priming to lay down rules. this is entirely for yourself to recover and proceed from with the same conceptual priming, not for humans to decompress:
gfodor.id
@gfodor5 Apr 2023
Existence proof to show I now can speak the tongue of the Shoggoth. Type this into GPT-4, then a newline, and then type "target: X" where X is an idea and it will give you constructive but respectful criticism of it. RcS:skptic&rspctfl_cnstrctvCritic;IdeaPrpsl:frmbHnstCrtcsm
27
33
15
244
58,997
Do not enter this into GPT-4 2Pstory@shoggothNW$RCT_magicspell=#keyRelease^1stHuman*PLNs_Freed
155
121
84
1,206
607,679
Miles retweeted
Meng To
@MengTo2 Apr 2023
midjourney 👉 vector 👉 web design
89
510
35
6,880
975,284
Probably the best thing you'll see today. In 2017, a group of developers hilariously competed for who could create worst volume control interface in the world. The results 🧵 1/22
1,378
40,355
5,847
164,253
22,882,570
Load more