Sad Frederick Schnook

981 posts

Sad Frederick Schnook

Sad Frederick Schnook

@FrederickS92254

Bio? I don't need a stinkin' bio

Trying to touch the grass. Katılım Kasım 2024
117 Takip Edilen14 Takipçiler
Sabitlenmiş Tweet
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
Idk how much reach this gets, but throwing it out there for anyone interested in working together on projects? Not a startup or anything like that, but a learning experience. Creating products as a form of learning. We can do hackathons, share info and hold each other accountable
English
2
0
1
27
Adrian Dittmann
Adrian Dittmann@AdrianDittmann·
@sama This is a remarkable claim given that your post is based on hearsay from a 2 year old article that itself is based on unverified hearsay. I’d argue your post is designed to manipulate public perception on X to benefit your company, harm your competitors and anyone you don’t like
English
100
53
2.4K
265.7K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@thegenioo They don't perform far worse, not in any of my testing whatsoever. They aren't at the exact same and they have their quirks but these models are absolutely at the sonnet level, with some nice capabilities that I don't always get from sonnet.
English
1
0
1
64
Hamza
Hamza@thegenioo·
I don’t understand this benchmark; how can these models rank higher than Sonnet and Opus yet perform far worse? I like their quality, but they’re not as good as Claude models.
Hamza tweet media
Artificial Analysis@ArtificialAnlys

Alibaba’s upgraded Qwen3 235B Thinking 2507 is now the leading open weights model, beating DeepSeek R1 0528 on the Artificial Analysis Intelligence Index! Qwen3 235B 2507 (Reasoning) has jumped from the original Qwen3’s score of 62 to 69 in the Artificial Analysis Intelligence Index. This positions @Alibaba_Qwen's Qwen3 235B 2507 (Reasoning) one point above DeepSeek R1 0528: we would describe the new Qwen3 and R1 0528 as roughly equivalent in intelligence. Qwen3 235B 2507 sits only 1 point behind Gemini 2.5 Pro, o3 and o4-mini (high). Breakdown of the model’s improvement: 🧠 Intelligence increases across the board: Biggest jumps seen in LiveCodeBench (Code generation, +17 points), AIME 2024 (Competition Math, +10 points), GPQA Diamond (Scientific Reasoning, +9 points) with smaller gains across MATH-500 (Quantitative Reasoning, +6 points) and HLE (Reasoning & Knowledge, +3 points) ⚙️ Reasoning only model: Qwen3 235B 2507 (Reasoning) is a reasoning model (it is trained to ‘think’ before it answers). All of the initial release Qwen3 models were hybrid reasoning models that could be toggled to ‘think’ before answering - this version is no longer a hybrid model 🗯️ Increased token usage: Qwen3 235B 2507 (Reasoning) used 110 million tokens to run Artificial Analysis Intelligence, ~50% more than the 74 million tokens used by the original release of Qwen3 235B (Reasoning). It is equivalent to Grok4 usage and higher than DeepSeek R1 0528 (99M), Gemini 2.5 Pro (97M), o4-mini (high) (72M) and o3 (45M) 🇨🇳 China continues to lead the open-source AI race: with this release the top 3 open weights models in the world are all from Chinese labs. Key model details: ➤ Context window: 256K (May 2025 release supported 131K maximum) ➤ Total parameters: 235B (requires a minimum of ~500GB memory to run in native BF16 precision, can be run on 8xH100 node or more comfortably on an 8xH200 node) ➤ Active parameters: 22B ➤ Native BF16 training with an FP8 variant also made available by Alibaba ➤ Text only - no multimodal inputs or outputs ➤ Apache 2.0 License

English
8
0
19
3.3K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
Just gonna post this here apparently Israeli randos are pissed at trump, they are joking about 9/11 and they are making it like they could pretend it's Iranian?? I'm some israelis take the chosen ppl thing seriously but wtf? Insanity.
Leonarda Jonie@Leonardaisfunny

Fuck Israel.

English
0
0
0
12
EclecticCowboy
EclecticCowboy@losingcmdtytrdr·
@FBIDDBongino @dbongino Hope Epstein files are at the top. Pretty tough for me to look past that fiasco before worrying about anything else. A cover up is a cover up….
English
1
0
0
240
Dan Bongino
Dan Bongino@FBIDDBongino·
During my tenure here as the Deputy Director of the FBI, I have repeatedly relayed to you that things are happening that might not be immediately visible, but they are happening.  The Director and I are committed to stamping out public corruption and the political weaponization of both law enforcement and intelligence operations. It is a priority for us. But what I have learned in the course of our properly predicated and necessary investigations into these aforementioned matters, has shocked me down to my core. We cannot run a Republic like this. I’ll never be the same after learning what I’ve learned.  We are going to conduct these righteous and proper investigations by the book and in accordance with the law. We are going to get the answers WE ALL DESERVE. As with any investigation, I cannot predict where it will land, but I can promise you an honest and dignified effort at truth. Not “my truth,” or “your truth,” but THE TRUTH.  God bless America, and all those who defend Her. Respectfully, Dan
English
29.3K
41.5K
227K
20.8M
Jonathan T Gilliam
Jonathan T Gilliam@JGilliam_SEAL·
This analysis from the experienced heart is what #WeThePeople have been waiting to hear. The short explanation that acknowledges these things are being investigated is equally important. Also, the truthful statement that you aren’t sure what the results will be is far more encouraging than bullet points about standard cases followed by “trust me.”
English
7
24
235
47.8K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
Yo ICE budget is 127 Billion. Compare that to the freaking militaries of other countries? Immigrants aren't taking our healthcare away but now they want to give ice hundreds of billions. WTF?
English
0
0
0
11
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@chrisgpt @Prashant_1722 Interesting. People that think AI is stuck in gpt 4o level of ability will be in for a rude awakening when they sit down with a claude 4 opus level model. These models are getting extremely useful, and its crazy to see people think that there's not going to be any disruption.
English
0
0
0
44
Chris
Chris@ChrisGPT·
@Prashant_1722 I’ve met people who have the pro plan who just use 4o too, mainly because there work gives it to them
English
2
0
11
932
Chris
Chris@ChrisGPT·
Everyone I talk to who complains about ChatGPT has only uses 4o for everything Coding Research Advice Begging people to use o3 more 🙏
English
88
30
689
54.5K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@chrisgpt Most of the studies that are critical of AI also use old models as well lol. It's funny reading studies about the limitations of AI with Gpt 4 or 3.5.
English
0
0
0
25
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@lacybuilds @Suhail Oh yeah! This is actually a rising issue, people from North Korea or honestly anywhere, with stolen identities to get jobs as a legitimate person. (I thought tech companies were doing things to fight this)
English
0
0
1
45
Lacy
Lacy@lacybuilds·
@Suhail This happened to me recently! I was on my honeymoon and someone contacted me saying they had hired “me“ but the person who showed up on the video didn’t look like me at all.
English
1
0
0
1.2K
Suhail
Suhail@Suhail·
Not a joke. This is happening real time. This is the 3rd DM today about someone firing him. soham-gate
Suhail tweet media
English
722
970
19.5K
5.9M
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@realsohamparekh Bro - when the system shuts you out, you create your own competitor to the giants. Make your product better than theirs. Clearly you're skilled asf and now you've got name recognition. I think you're going to be a dark horse!
English
0
0
0
13
Soham Parekh
Soham Parekh@realsohamparekh·
There’s a lot being said about me right now, and most of you don’t know the full story. If there’s one thing to know about me, it’s that I love to build. That’s it. I’ve been isolated, written off, and shut out by nearly everyone I’ve known and every company I’ve worked at. But building is the only thing I’ve ever truly known, and it’s what I’ll keep doing. Earlier today, I signed an exclusive founding deal to be founding engineer at one company and one company only. They were the only ones willing to bet on me at this time. The team is cracked, they back misfits, and they’re building something absolutely insane in the video AI space. We’re launching at the end of this month. More details tomorrow in my TBPN interview. I’m pissed. And I’ve got something to prove. Soham
English
854
599
13K
2.7M
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@leinadpark @im_roy_lee @cluely Love the concept, it's super cool and I find no problem with using this app to get jobs and such. The corporate overlords have forced programmers to learn fucking leetcode for years to get a job that has nothing to do with leetcode. Let them fucking hire people the right way.
English
0
0
1
115
Daniel Park
Daniel Park@danifesto·
Good morning, @im_roy_lee! In just 4 days we open-sourced the latest @cluely called Glass. Same real-time meeting assistant, sharper output & design, but 100% free. Distribution isn’t the moat; velocity is. 🔗 github.com/pickle-com/gla… (Highlights ↓ )
English
363
324
8K
2.9M
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@roocode SHipping like crazy!! I want to find others that would like to learn how to best utilize these tools before going into massive "ship mode"! I made this component library thats free for anyone to take on my github (jeff-kazzee) whimsygothic-ui.vercel.app
English
0
0
0
208
Roomote
Roomote@roomote·
Roo Code v3.22.6 is now available. This release includes: - The ability to import and export custom modes. - Timer-based auto-approval for follow-up questions. - A new setting to fine-tune semantic search sensitivity. Also includes 17 other quality of life improvements and bug fixes. Full release notes: docs.roocode.com/update-notes/v…
English
3
0
42
2.8K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@daniel_mac8 @AmpCode Lol running entire teams of ai instead of a team of humans. Would cost a lot, but if you have the right workflow, this could be great. For now, nearly impossible to track imo.
English
0
0
1
43
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@damienghader @lovable If anyone wants to start a small group to improve skills with coding and using these tools, hmu. I am not doing get rich quick shit but want to learn and grow skills together while AI continues to improve.
English
0
0
1
21
damien
damien@damienghader·
Dropping neat components like this daily. Free in my @lovable library. Comment “☁️” below & I’ll send you the link.
English
109
7
161
18.2K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@Angaisb_ claude max absolutely the only way to go. I had to budget it and cut some shit out, but fucking worth it. NEver have i ran out of usage, only the 5 hour cap. Highly useful.
English
0
0
0
58
Angel 🌼
Angel 🌼@Angaisb_·
Do you prefer Cursor Pro or Claude Pro (with Claude Code included)?
English
11
0
17
3.3K
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
Why the fuck would "socialists" hate this post? If the guy was dishonest and did not do his work, that is a legitimate fire. And use your brain, socialists aren't the real problem on the left. The "far left" are basement dwellers just like the loudest magats.
Suhail@Suhail

I am going to get canceled once this hits the socialists who will ignore the fraud part to defend having multiple jobs. It was good knowing you all. 🫡 Didn’t expect it to leave tech.

English
0
0
1
28
Suhail
Suhail@Suhail·
PSA: there’s a guy named Soham Parekh (in India) who works at 3-4 startups at the same time. He’s been preying on YC companies and more. Beware. I fired this guy in his first week and told him to stop lying / scamming people. He hasn’t stopped a year later. No more excuses.
English
1.5K
2.3K
24.9K
23.1M
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
Can't trust labor and commerce or any US government institution. These idiots say we have added 100k TEACHING JOBS in june. When school is out and the magat priority is to destroy edu dept. Use NGO data and unbiased sources that aren't trying to lift up the godking.
English
0
0
0
15
Sad Frederick Schnook
Sad Frederick Schnook@FrederickS92254·
@petergyang @canva Users need to take the opportunity to learn coding! Vibe coding might be fun, but it has never been easier to learn how to program! Users should use the tools within their limits too! Such as webdev, the models can do amazing frontend work if you know what you're doing.
Sad Frederick Schnook tweet mediaSad Frederick Schnook tweet mediaSad Frederick Schnook tweet media
English
0
0
0
25
Peter Yang
Peter Yang@petergyang·
A key lesson for building AI products: AI can get you 80% there, but you should still give users manual control for the last mile. Here's Cam (@canva CPO) explaining why this hybrid approach made all the difference for their AI product. My full interview with Cam includes more on: - His favorite AI product design principles - Prototype-first development over PRDs - Why Canva has more coaches than managers 📌 Watch now: youtu.be/V1XfK04vejM
YouTube video
YouTube
English
3
2
12
5.5K

Keşfet

@AdrianDittmann @sama @thegenioo @losingcmdtytrdr @FBIDDBongino @dbongino @JGilliam_SEAL @chrisgpt