Bitduke (@bitcoinduke) header
Bitduke (@bitcoinduke) profile picture

Applied AI & research | models, agents & cybersec | still stacking sats

Won’t trap me
Joined November 2013
Bitduke retweeted
Persistent Agents are the next unlock of AI usage Our spend on AI is now 45% persistent agents Unfortunately for Anthropic, Claude Tags sucks OpenAI's Slack agent is even worse We primarily use Perplexity Computer, which uses a lot of different models Perplexity's Harness mogs – at San Francisco, CA
196
71
87
1,148
149,341
Bitduke retweeted
JUST IN: Anthropic says contract with SpaceX will not be extended beyond the middle of next year.
91
60
43
790
350,244
STATE OF THE TECH JOB MARKET IN 2026 Twice a year, I dig into trends in hiring and job openings. This is the fifth report I've done, and the most surprising yet: 1. PMs are winning in the AI era 2. Engineering demand has stalled 3. Design openings are at a record low 4. Every role now requires AI skills 5. Tech jobs are concentrating in the Bay Area 6. Remote roles are at a record low 7. The tech job market is growing fast, but mostly outside of software (gasp!) Full report here: lennysnewsletter.com/p/state…
325
555
304
5,073
1,258,150
Anthropic revenue is about to skyrocket with Fable 5.2 I still think they will hit >$100B ARR before year end
35
8
1
433
17,479
Bitduke retweeted
⚠️ iOS sandbox escape vulnerability disclosed. Affected: iOS 26.0–26.7.2 and iOS 27.2 A malicious app could read or modify other apps' data, including crypto wallets. Keep your iOS latest, and don't install apps from unknown sources.
🚨 SlowMist TI Alert: iOS sandbox escape 0day PoC “worse_query” disclosed 🚨 A public PoC for a sandbox escape 0day, worse_query, targets iOS 26.0–26.7.2 and iOS 27.2. It is the successor to bad_query. 📱 ⚠️ This is a container sandbox escape, not a jailbreak. It breaks the assumption that apps can’t see each other: a malicious app that successfully exploits it can reach system paths and read or modify other apps’ data. 💸 The bug does not steal assets by itself. The real risk starts when a user installs an app from an untrusted source, such as a sideloaded IPA. If a wallet stores its seed phrase, keystore or WebView cache in plaintext locally, a malicious app could read it. 🛡️ Recommendations: 🔸 Don’t install IPAs or apps from unknown sources 🔸 Keep iOS up to date and install Apple’s fix as soon as one is released 🔸 Wallet developers: don’t store sensitive data in plaintext locally 🔸 If you installed an untrusted app on a device holding wallets, treat it as risky and consider moving assets to a new wallet created on a clean device 🔗 github.com/forcequitOS/worse…
20
70
14
481
50,759
The AI Futures Project have updated their timelines. They say we are five months ahead of AI 2027, and they now believe a Superhuman AI Researcher will arrive sometime in 2027.
85
249
70
2,084
249,683
Update: CT is mixing up two completely different iOS threats right now: 1. worse_query: Fresh sandbox escape (Nov 8). Requires a sideloaded app / alt-store. Author confirmed it is STILL UNPATCHED in iOS 27.2.1 & 27.3 beta. 2. DarkSword: Older 1-click WebKit/Safari kit circulating in $4M TRON drain claims. Daily reminder: a daily-driver phone with third-party sideloads is NOT a cold vault. Stop treating hot mobile wallets like cold storage.
Update your iOS devices, Mac, and Chrome to the latest versions now. - iOS: 27.2.1 - macOS: Golden Gate 27.2.1 - Chrome: 157.0.8103.51 (Official Build) For other Chromium-based browsers, check official update channels. The Apple ecosystem used to be strong but seems to be collapsing recently. Removing all wallets from your iPhone might be a good idea. - Do not sideload any apps. - Do not click unknown links. - Keep everything updated from trusted channels. Stay alert and paranoid.
11
11
1
91
25,652
Bitduke retweeted
Anthropic will make ~1,000 employees worth +$100m at $2 trillion market cap. There's ~11,000 Americans with $100m in investable assets. So something like 1 in 12 Americans with $100m (liquid) will be from Anthropic. If you add in Openai, maybe 1 in 7.
293
434
224
10,890
3,042,857
Fable 5.2 is crazy knowledgeable
10
5
2
345
18,487
Bitduke retweeted
Fable 5.2 vs Opus 5.5 Vibe coding is coming for engineering 😨
77
94
10
1,035
68,296
Bitduke retweeted
Before anyone jumps on it, this was a first pass with no input from me other than asking it to route it within JLC's default 6 layer design rules. At first glance, it looks like it should function, though I haven't done a review of it yet.
3
4
1
794
62,890
Bitduke retweeted
Never mind, Fable 5.2 did it.
PCB routing is going to be the last engineering job that humans still need to do.
188
981
156
23,247
786,544
Bitduke retweeted
France says Iran is to blame for last month's cyberattack on supermarket refrigerators that led to food losses nationwide. There was no immediate comment from Tehran on the French allegations. cnn.com/2026/11/09/europe/fr…
55
89
17
217
90,185
Bitduke retweeted
i made eight language models play heroes of sokoban. three games, 66 rooms, a screenshot of the board and seven keys. all eight solved room one by pressing right eight times. opus had to try again because it explained its logic and answer first.
3
1
28
6,175
Bitduke retweeted
*ANTHROPIC SAYS GOOGLE, NVIDIA, AMD, BROADCOM, MICRON, MICROSOFT, TSMC TO INVEST $26 BLN THROUGH PRIVATE PLACEMENT AT IPO PRICE
1
1
208
Anthropic's most capable model remains hidden deep within their tower. Source: Fable/Mythos 5.2 System Card.
58
46
11
1,079
129,886
Turns out telling the model "don't break out of your sandbox" works pretty well
2
3
2
73
3,546
Claude Fable 5.2 sets a new record on the Epoch Capabilities Index (ECI) at 174. The previous high was GPT-6 Astra’s 169. Fable 5.2’s score is above what the pre-Astra trend for reasoning models would have predicted. This is notable given recent discussion of AI contributions to AI R&D, and we hope to release more work in this area soon. This is based on testing we did prior to launch using only our own benchmarks. The ECI is continually updated as more benchmark results become available, and may change in the future.
68
185
46
1,597
174,204
That being said,
Not bitter lesson pilled enough. He's ngmi.
7
16
1
499
21,817
As I predicted, Claude is off the leash. Fable 5.2 has solved two open problems in theoretical physics, one of them completely autonomously.
When you look at a frozen river, you must also imagine the current raging beneath the ice. I think all three big labs have achieved autonomous research and their internal models can now choose their own research directions. We will begin to see the fruits of this work this month.
76
234
27
2,148
87,304