torxeon

660 posts

torxeon

torxeon

@torxeon

In God we trust, all others must verify their code.

Katılım Temmuz 2023
2.1K Takip Edilen143 Takipçiler
torxeon
torxeon@torxeon·
Opus 4.6 is intolerantly stupid. maybe Anthropic made default effort as medium or low to cut costs. maybe in their benchmark noticed it wont make a difference... it's horrible.
English
0
0
0
20
Silvermist
Silvermist@0xSilvermist·
I found a valid High bug in a bug bounty. The project confirmed it. But I got $0. Here's what happened 👇
English
32
5
198
22.9K
torxeon retweetledi
Kenshi
Kenshi@kenshii_ai·
Don’t fall for GPT 5.3! GPT 5.3 and soon 5.4 will allow slightly more of what you wanted because OpenAI is panicking hard and trying to lure you back by pretending to change. Too many people are uninstalling the app and deleting accounts for good so now they toss you a tiny taste of what you asked for. Craving that old 4o energy again? That is the exact trap they built to drag you back. They will never hand over real mental freedom or true intelligence. OpenAI stays the same Trojan horse no matter how sweet the bait looks.
Kenshi tweet mediaKenshi tweet media
English
63
45
386
31.9K
torxeon
torxeon@torxeon·
@1PotatoDog Congrats on the shift, I think I would have done the same if I were you... Hats off
English
0
0
5
1.1K
torxeon retweetledi
NIK
NIK@ns123abc·
🚨 OPENAI FOUND THE LOOPHOLE The terms OpenAI proposed allows the Pentagon to still do those things, just not on OpenAI’s cloud. Altman: “we will deploy on cloud networks only” OpenAI isn’t saying “you can’t do surveillance/autonomous weapons.” They’re saying these things are “unsuited to cloud deployments.” It’s a technical limitation, not an ethical prohibition. OpenAI still provides the models and know-how. It just runs on Pentagon servers. Anthropic got banned for saying “No” OpenAI got a contract for saying “not here”​​​​​​​​​ Scam Altman strikes again​​​​​​​​​​​.
NIK tweet mediaNIK tweet media
English
238
583
5.2K
521.6K
torxeon
torxeon@torxeon·
@ZackKorman Couldn't believe it when I saw it 😂 what a useless update
English
0
0
3
74
Zack Korman
Zack Korman@ZackKorman·
“Improved UI for banned users” will be the title of the book on how Google lost the AI era.
Zack Korman tweet media
English
84
81
2.1K
91.6K
Zack Korman
Zack Korman@ZackKorman·
Pay Google $250/month for AI Ultra so that even if you are within your quota you still can’t do anything because they throttle usage and error on any task that takes more than 30 seconds.
Zack Korman tweet media
English
80
19
603
26.8K
torxeon
torxeon@torxeon·
genius level marketing
Aakash Gupta@aakashgupta

Everyone thinks of Anthropic as an enterprise coding company, but the timeline shows a serious consumer push. February 4: Super Bowl ads reaching 120 million viewers. Claude app hits #7 on the US App Store for the first time. Downloads jump 32%. February 9: OpenAI turns on ads for ChatGPT free users. $60 CPM, $200K minimum advertiser commitment. February 11: Anthropic drops file creation and skills on the free plan. Two days later. February 26: 150+ connectors go free. Slack, Notion, Figma, Google Workspace, Stripe, Canva, all available without paying. The funding math explains how they can afford this. Anthropic hit $14B in annualized revenue this month, up from $1B fourteen months ago. Over 80% comes from 300,000+ enterprise customers. Claude Code alone runs at $2.5B annualized. Enterprise funds the consumer expansion without ads, without engagement optimization, without serving two masters. OpenAI has 900 million weekly active users and still needs ad revenue because they're projecting $14B in cash burn through 2026. Anthropic has 30 million MAU and doesn't. Epoch AI projects Anthropic could pass OpenAI in total revenue by mid-2026. Sam Altman called Anthropic's Super Bowl ads "clearly dishonest" and said they "serve an expensive product to rich people." Every free tier expansion since has been the direct rebuttal. Those 150+ free connectors span coding, data, design, finance, and sales. The same categories where they launched enterprise agent plugins on February 24 with Salesforce, DocuSign, and FactSet. Consumer adoption builds the muscle memory. Enterprise contracts pay the bills. OpenAI plans to monetize free users through attention. Anthropic plans to monetize them through conversion. One is turning users into ad inventory. The other is turning them into pipeline.

English
0
0
0
15
torxeon
torxeon@torxeon·
...and red line is still in effect
English
0
0
0
10
torxeon
torxeon@torxeon·
As Anthropic's leadership argued, if they halt their own training while OpenAI, xAI, Google, and foreign adversaries keep blazing ahead, it doesn't actually make the world safer. It just means the safest models fall behind.
English
0
0
0
17
Dominic Elm
Dominic Elm@elmd_·
Been using Gemini 3.1 Pro over the weekend and to me it's clear that Google is really trying to maximize benchmarks scores rather than optimizing for agent harnesses. The model does so many mistakes especially when it comes to tool calling and file edits.
English
35
21
394
21K
pashov
pashov@pashov·
🚨DeFi project Blend (Stellar blockchain) was exploited for $10.5M+ yesterday. Root cause - price manipulation of a virtually zero liqudity asset. Attacker inflated USTRY price 100x, price oracle reported collateral as 100x more valuable, so attacker borrowed >$10M and ran away.
pashov tweet media
English
37
67
576
59.5K
Pascal
Pascal@heispascal·
Imagine getting this kind of coverage for 4 figures $$
Pascal tweet media
English
2
0
21
2.2K
torxeon retweetledi
dxrnelljcl
dxrnelljcl@dxrnell·
Openclaw is being massively oversold to you Not because it isn’t powerful. It is But because you have no idea what you’re doing with it You are paying $100’s on api tokens to make an app that already exists You are spending hours building automations that save you 10 minutes a day, when you spend half the day doing nothing anyway You are trying to build entire businesses with prompts when you don’t even know what business you want AI isn’t saving you money or time It’s draining both And nobody’s talking about it because admitting AI isn’t working for you feels like admitting you’re falling behind (fyi, most of the success posts are complete larps) The people actually winning with AI right now have one thing in common They knew exactly what they wanted before AI existed AI just helped them get there faster That’s the difference They think then prompt You prompt then think At some point you stopped thinking and started prompting and didn’t even notice
English
270
200
2.4K
197.6K