pandora

1.6K posts

pandora banner
pandora

pandora

@Markste_in

@[email protected] # Our existence is but a brief crack of light between two eternities of darkness. 🏳️‍🌈🇩🇪 Team DB ;)

Germany Katılım Mart 2018
520 Takip Edilen192 Takipçiler
Michael | Hypermarkets
Michael | Hypermarkets@itsmichaelluu·
Tom Lee says the May/June SPY crash will feel like hell, but you just buy the dip for SPY target $770+ Here's exactly what's coming from @fundstrat: 1/ The Timeline New Fed chair confirmation is starting NOW. Kevin Warsh. Senate floor vote could come 50-70 days after committee. That puts a new Fed chair in place by June-July 2026. 2/ History doesn't lie and it's horrifying "10 of 13 Fed chairs had a drawdown of more than 10% in the first year. So I think it's more the rule than the exception that we're going to get a drawdown." says Tom Lee Since 1930, the SPY has logged average drawdowns of 5%, 12%, and 16% over the 1-, 3-, and 6-month periods after a new Fed chief took the helm all larger than the typical peak-to-trough drop in a randomly selected year, per Barclays. This isn't a black swan. It's a pattern. 3/ The Road Map "We originally said that the market could get towards 7,300. And it looks like we're on a path towards that. We're well over 7,000. And then we'll have a decline that will feel like a bear market. Perhaps it's because of the market testing new Fed chair." — Tom Lee The data backing this: The median intra-year drawdown from an all-time high in the SPT since 1980 is 10.4%. Statistically, there's a 50% chance the index drops 10% or more at some point in 2026. Feel like a bear market = -15% to -20% easily on the table. 4/ But Mag 7, crypto, and software are ALREADY bleeding "Keep in mind, a lot of the markets have already had a drawdown in the Mag 7, crypto, and software."— Tom Lee The broad index hasn't felt it yet. Rotation is a warning shot, not a safety net. 5/ Then… the rally of a lifetime "I think a rally follows because fundamentals are strengthening maybe one of the strongest rallies we'll see in our lifetime."— Tom Lee Since 1980, the average intra-year drawdown has been over 14% yet the SPY has still averaged 10.7% annual gains during that same period. Double-digit drops have historically come with double-digit annual gains. The flush is the setup. 6/ His final warning: "Just a reminder just don't time the market, even if you're tempted to." says Tom Lee So this is the full sequence:SPY → 7,300 than New Fed chair → 10-16% drawdown then rally of our lifetime
Michael | Hypermarkets tweet media
English
66
150
797
115.9K
Gargoyle
Gargoyle@degargoyle·
THE S&P 500 HAS A PATTERN NOBODY TALKS ABOUT Every crash since 2020 found its bottom at Fibonacci 0.5 - not approximately, exactly Three times this played out. Three times people called it a coincidence. It's not a coincidence. It's structure Remember it!
English
19
26
199
66.6K
pandora
pandora@Markste_in·
@above_spec The hallucination rate is still bad and make it close to unusable for everything that is not grounded (like coding with tests)
English
0
0
1
129
AboveSpec
AboveSpec@above_spec·
"You need a 24 GB GPU for serious local LLMs in 2026." Everyone repeats this. It's not true anymore. Just ran a 35B-parameter model on an RTX 4060 Ti 8 GB: • 41 tok/s at 16k context • 24 tok/s at 200k context Recipe + benchmarks below 🧵
AboveSpec tweet media
English
135
232
2.8K
282.2K
Mario Nawfal
Mario Nawfal@MarioNawfal·
🇺🇸🇮🇷🇵🇰 Pakistan looks like it's preparing for U.S-Iran peace talks. Two U.S. C-17 Globemaster heavy airlift aircraft just landed at Noor Khan Airbase near Islamabad. Roads from the airport are locked down, the Serena and Marriott hotels have cleared guests and are taking no bookings until Friday. The U.S. and Iran held their first round of negotiations in Islamabad on April 11. Pakistani sources say a second round is likely this week. Source: Al Jazeera
Mario Nawfal tweet mediaMario Nawfal tweet media
Mario Nawfal@MarioNawfal

🇮🇷🇪🇺 Iran to the EU: don't lecture us on international law while ignoring it yourselves. Tehran fired back after Brussels demanded free passage through the Strait of Hormuz.

English
144
371
2.2K
329K
pandora
pandora@Markste_in·
@NathieVR But what’s the use case here?
English
0
0
0
17
Nathie 🔜 AWE
Nathie 🔜 AWE@NathieVR·
Using Apple Vision Pro to draw and interact in 3D feels like cheating reality.
English
140
306
5.3K
2.9M
Kav
Kav@TheOtherKav·
Ok, here's something I want to vent about: @Microsoft how many more fucking times am I going to see the screen‽ I've had this computer for years, and you keep wanting me to "set it up" again after updates I don't want M365 I don't want Office I don't want OneDrive I don't want to have to have a Microsoft account just to log into my computer Let my computer just be *MY* damn computer. Because at this rate, as soon as SteamOS works with Intel and Nvidia, I'm ditching windows on all my machines. This is ridiculous. My computer was already set up the first time I set it up when it was still on Windows 10 a few years ago.
Kav tweet media
English
476
306
4.3K
126.9K
pandora
pandora@Markste_in·
@AnthropicAI what is going on here... constantly getting 529 {"type":"error","error":{"type":"overloaded_error","message":"Overloaded"} Please scale faster!
English
0
0
0
51
Yash
Yash@yashhq_22·
$20 in codex feel endless. $20 in claude disappear in 20 mins.
English
184
45
1.8K
78.2K
pandora
pandora@Markste_in·
I'm only on the $20 claude plan and tried to do a review of one of my repos with Opus on medium thinking. After less than 15 Minutes I ran out of token! The prompt didn't even manage to finish. The same prompt on codex, kiro or gh copilot just uses a few percentage! @claudeai
English
0
0
0
84
pandora
pandora@Markste_in·
@EIiasNajarro There is a bug. It shows one less type than selected. If u select 3 types it only shows two colors (types)
English
1
0
1
208
pandora
pandora@Markste_in·
@karpathy @shancn7 @maxbittker Unfortunately I can get codex to loop forever . It stops after a few minutes. OpenCode and Claude Code works the most reliable for infinite looping (so I am using codex/openai through OpenCode rn)
English
0
0
1
399
max
max@maxbittker·
From @karpathy's autoresearch .md
max tweet media
English
50
122
3K
223.1K
pandora
pandora@Markste_in·
@DeryaTR_ @Cisco_research I already canceled… that feels like a paid trial and not like a valuable service u buy rn
English
0
0
1
74
Derya Unutmaz, MD
Derya Unutmaz, MD@DeryaTR_·
After 15 min of asking Claude Code Opus 4.6 to continue where it left off. which had hit the limit, 90% of the limit was already used. A few minutes later, it was at 100%. Result: zero tasks completed! This is as bad as it gets. Clearly, the token-eater bug is still not fixed!
Derya Unutmaz, MD tweet media
English
194
73
1.7K
78K
pandora
pandora@Markste_in·
@manoj_ahi I have the exact same issue. A few kinda small prompts and suddenly my Quito was almost gone… like wtf @claudeai
English
0
0
0
62
Manoj Ahirwar
Manoj Ahirwar@manoj_ahi·
usage limit over in 20 mins!!! did something change in claude code? I am on Max plan
Manoj Ahirwar tweet media
English
250
69
2.4K
316.8K
Bosco
Bosco@BoscoMac·
I set up the first devices in the new laboratory. The electronics laboratory is now fully air-conditioned. Stable temperatures for calibration. Soon the work tables will be set up. #hp #agilent #keysight #tektronix #rohdeubdschwarz
Bosco tweet mediaBosco tweet mediaBosco tweet mediaBosco tweet media
English
31
25
549
21.9K
Ryan Leachman
Ryan Leachman@RG_Leachman·
I asked Claude to build my daughter an app that plugs into our piano, can read live key strokes, can show her sheet notes and key view and ends with a Guitar Hero style game. All while giving progressively harder songs. Today she’s using It and crushing It.
English
594
1.6K
24.8K
3.4M
Ihtesham Ali
Ihtesham Ali@ihtesham2005·
A PhD student at Oxford got caught submitting "AI-generated" work. Except he hadn't used AI to write anything. He used it to think. Here's the workflow his advisor called "the most sophisticated research process I've seen in 20 years." He starts every essay with a brutal diagnostic prompt. Dumps his rough argument into Claude and asks: "What are the 3 weakest logical jumps in this reasoning? Where would a hostile examiner attack first?" The AI doesn't write his essay. It destroys his draft. Then he rebuilds. But the next step is what separates him from every other student using ChatGPT or Claude to generate paragraphs. He uploads the top 5 papers in his field and asks: "What claims in my argument contradict or oversimplify what these authors actually found?" Most students cite papers they've skimmed. He cites papers he's been forced to genuinely understand. The final move is almost unfair. Before submitting, he pastes his conclusion and asks: "What would a philosopher of science say is missing from this argument? What assumptions am I making that I haven't defended?" His essays come back with comments like "unusually rigorous" and "demonstrates rare critical depth." He's not using AI to write. He's using it to think harder than he could alone. The tool hasn't changed. The workflow has.
Ihtesham Ali tweet media
English
195
584
3.7K
1.1M
pandora
pandora@Markste_in·
@claudeai I have to admit that I don't feel a change and I am just using Sonnet mostly... It's a Sunday so it should be apply today??
English
0
0
0
104
Claude
Claude@claudeai·
A small thank you to everyone using Claude: We’re doubling usage outside our peak hours for the next two weeks.
English
1.9K
3.5K
48.3K
12.7M
pandora
pandora@Markste_in·
@bulatov_org Wow that looks amazing! How did u do that? Python / matplotlib?
English
1
0
0
12
Vladimir Bulatov
Vladimir Bulatov@bulatov_org·
Animation of Simone Conradi attractor
English
2
2
40
3.3K