Florian Dedov (@NeuralNine) header
Florian Dedov (@NeuralNine) profile picture

Florian Dedov

@NeuralNine

🧠 AI/ML Engineer 📹 YouTube (485K+ Subs): https://nitter.cf/t.co/AdBzHLggKb 🐙 GitHub: https://nitter.cf/t.co/5D67i6dB6W

 Vienna, Austria
neuralnine.com/
Joined April 2019
  • Tweets1.2K
  • Following222
  • Followers2.5K
196Media
Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)Photo by Florian Dedov (@NeuralNine)
I was able to meet Jensen Huang the CEO of @nvidia at #GTCParis. This guy is a legend!
Image from Florian Dedov (@NeuralNine)
1
63
3,123
I need some robots
90
If this pattern continues outside of benchmarks it would be a crazy comeback.
Demis Hassabis (@demishassabis) profile pictureDemis Hassabis
@demishassabisOct 29
No.1 💎
Image from Demis Hassabis (@demishassabis)
1
97
Also pretty sure they're not benchmaxxing. Let's go Europe! 🇪🇺🚀
Paul Couvert (@itsPaulAi) profile picturePaul Couvert
@itsPaulAiOct 29
Mistral is... BACK?! They've just released what might be the absolute best open weights model for coding - Better than Kimi/DeepSeek/GLM on terminal bench 4 - 1st on cognition's private frontiercode 2.0 - Fewer tokens per task than others Already available on Hugging Face and via api!
Image from Paul Couvert (@itsPaulAi)
1
2
197
If the whole cryptocurrency thing dies, for me it will be like it never happened. It's the one "tech hype" I never dabbled in even a little bit.
3
303
Doubling the usage limits on useless models is not the right strategy.
109
Cancelled my ChatGPT subscription. These models became completely useless since the DevDay.
21
5
174
4,452
Doctors, lawyers, accountants, mechanics... No matter what profession. Claude always sees something these "experts" miss. It's insane.
2
152
Google just crushing OpenAI right now...
Arena.ai (@arena) profile pictureArena.ai
@arenaOct 28
Exciting news: Gemini 4 Carbon (High) from @GoogleDeepMind just made a huge debut on Arena, landing #1 in Text Arena, #2 in Agent Arena, and #4 in Code Arena: WebDev. Back at the top, Gemini 4 Carbon delivers frontier intelligence and reshapes the Pareto frontier across three arenas. In Agent Arena, Gemini 4 Carbon ranks #2 with a +16.2% net improvement across +21.9K real-world agentic sessions from our global community, surpassing top models GPT-6 Astra and Fable-5.1. At $1.80 median cost per task, Gemini 4 Carbon (High) costs 72% less than the #1 ranked Claude Opus 5.5 (Max). Its standouts by key signals: - #1 Steerability (+18.6%): can the model course-correct when you push back? - #2 Praise vs Complaint (+37.9%): implicit sentiment in users' reactions - #3 Confirmed Success (+18.7%): explicit “yes, that worked” from users In Code Arena: WebDev, Carbon scored 1768 pts landing #4: - Just 2 pts below #3 Claude Sonnet 5.5 (xHigh), at 1770 pts - 89 pts above #11 Gemini 4 Argon (High), at 1679 pts By category, it ranks #3 in Gaming and Simulations, and #5 in Brand & Marketing. In Text Arena, Gemini 4 Carbon (High) is the new #1 with 1528 pts, now ahead of Gemini 4 Argon (High) at #2 with 1520 pts. Congrats to the @GoogleDeepMind team on this huge milestone!
Image from Arena.ai (@arena)
Image from Arena.ai (@arena)
Image from Arena.ai (@arena)
1
127
Nice stuff from Google
Image from Florian Dedov (@NeuralNine)
1
135
Artificial Analysis updated their intelligence benchmark. Some models moved down as a result. American models seem to remain pretty solid, while Chinese ones rank lower now. Is this benchmaxxing?
Image from Florian Dedov (@NeuralNine)
2
154
Every bug fix used to mean learning multiple new things. How often do we actually read error messages and tracebacks nowadays?
1
1
4
541
Upper right should maybe be rephrased as: "You can easily pivot"
Florian Dedov (@NeuralNine) profile pictureFlorian Dedov
@NeuralNineOct 28
I think one of the worst decisions you can make now is to stop learning because of AI. Game theoretically:
Image from Florian Dedov (@NeuralNine)
1
3
200
I think a large part of future SaaS businesses will be about providing interfaces for agents to access whatever you value proposition is. Data, network effects, customer support...
1
2
142
I think one of the worst decisions you can make now is to stop learning because of AI. Game theoretically:
Image from Florian Dedov (@NeuralNine)
1
1
10
494
I think I am still not realizing fully how much I will miss writing code manually
5
159
I think people overestimate how many people would actually build their own custom software. Let's say it would take 10-20 hours of prompting to get an Excel clone. Do you really think all the business and finance people would do it? I doubt it.
11
1
12
894
This is exactly true!
Nassim Nicholas Taleb (@nntaleb) profile pictureNassim Nicholas Taleb
@nntalebOct 24
The problem isn’t that AI is intelligent. It isn’t. The problem is that too many adults (but not all) are at once mechanistic, imitative, unimaginative, stupid, formulaic, influenceable, & gullible, arrogant beyond their skill set, & too expensive to maintain.
135
Florian Dedov retweeted
Mario Zechner (@badlogicgames) profile picture
Mario Zechner
@badlogicgamesOct 27
and it's not like the new kind of software is any good by the old standards (and yes, that was true for most old software as well). it just doesn't matter anymore. and it's not just software. anything machines can generate now goes this way. an endless stream of mediocrity.
9
5
5
282
233,195
I was wondering why all features except for one worked so well. Look what I found 🤣🤣🤣
Image from Florian Dedov (@NeuralNine)
1
178
Europe as a continent and ecosystem might not be in the best place economically and technologically. But don't use this as an excuse for yourself as an individual. As a European founder you still have numerous options and opportunities to maneuver. Don't take the black pill.
1
3
195
Load more