.Dot.

@BruceKnuteson @P_Remarks Ok this chart is wild if true. I know that 80% of stock appreciation is gap ups but what the actual fuck?

@P_Remarks Many of the people who think stock prices are "right" still don't know about the strikingly suspicious return patterns that can only be explained by the market manipulation described in ssrn.com/abstract=46190… because nobody has yet widely and honestly reported them.

I starting buying saas companies for the first time this year this week. I bought $ADBE $CRM and $WDAY. I don’t own it yet but I think I will also buy $NOW. I plan to develop these positions over Q3. I don’t believe the bottom is in but I do believe the immediate term chip top and saas bottom happens in Q3.

@BradPrick @michaeljburry It’s two trades. 1 shit trade with a 50% loss and a 2nd trade at 6x. But numbers wise yes you are right and he’s retarded.

@BradPrick @michaeljburry His logic is about sunk cost not opportunity cost.

A stock falls from 100 to 10. That's a 90% decline. The investor buys in at 10 because it is indeed worth 30. But it has fallen so much it has become detached from the underlying business and so falls to 5.. The value investor is down 50%. Which isn't as bad as 90%, but is still 50 effing percent. The value investor could have sold at 8, and limited the loss. 10 years later the stock is at $30. While said nvestor gies to medical school and finsihes residency to become a surgeon. That's a six times return in 10 years. About 20% per year. Wait, is it not only a triple? No, because the investor could have sold at 5 like just about everyone else did, but did not. What I describe here is almost exactly WBD. And it didn't take 10 years, because it usually doesn't.

He is saying you 6x’d your opportunity cost at a common puke point for most investors. Effectively the 50% loss is sunk cost. So the most important decision is the most recent one and the most recent one in this decision chain is a 6 bagger. Hope that helps. Burry is slightly retarded.

@michaeljburry How is that still not a triple? I don’t get it

@rubicon59 I read a bit of his thread. Yeah that game took 5 min to create vs 15 sec for web based. To do anything of value it would be terrible. 100% bait confirmed.

@rubicon59 Not likely anytime soon. (Years) you need an enormous amount of memory to have anything close to the t/s output from web based chat interfaces. I didn’t look in to that posts claims but I suspect it is largely bait.

How long before AI PCs is a booming market?

$XOM looking to buy woodside for their LNG assets is very telling on where the puck is heading as far as capturing volume is concerned.

Hot take - we won't have a recession until after 2027. CPI is going to stay sticky regardless of energy prices and energy prices in general is not what is driving sticky core inflation. Even the lag doesn't matter. Nominal GDP will remain quite high through 2027. Why? Because of AI capex. AI capex is causing demand side inflation due to what will amount to trillions of dollars entering the real economy. Also, what that means for stock is the real question. Those trillions aren't going into financial assets they are going into stuff.

@GavMcCracken You either realize 50% upside over the next year or 3 months. Either is fine with me.

Anas Alhaji thinks oil is oversold here he said quote "knee jerk reaction to the news". He also doesn't think we hit tank bottoms or get a huge price spike, taking a moderate view here. Jeff Currie: "Buckle your seatbelts, I'm long"

@GaelBreton ehhh... no - codex 2 and gpt x.x are coming out soon. Sam is better friends with the government. It'll be fine.

I feel like Anthropic broke the AI industry. Fable is so powerful that no competitor has an answer, so they don't release anything. Fable is banned because it's so powerful, so even if competitors have an answer, they want to see how it plays out first. We're basically in a stalemate. Nothing will happen until this is resolved, one way or another.

Fable 5 on claude code felt like a better version of codex. codex 2 and gpt 5.6/6.0 is just going to wipe the floor with claude code. Anthropic tuned their LLMs to make you feel like you are doing big things but in fact you are running around in circles overthinking wasting tokens getting nowhere. OMG Fable 5 one shot a shitty video game so much better than GPT 5.5 did! Who cares? Whoever freaks out about things like that isn't actually doing serious work on anything.

@sin_management The world has completely crumbled without fable 5 huh? Definitely codex

You can only choose one: - Cursor - Devin - Antigravity - VS Code - Codex - Claude Code

@junkbondinvest Not a problem - google underwrites all the private credit risk for anyone that wants in.

What happens to HY data center bonds if the labs never earn their cost of capital? junkbondinvestor.com/p/credit-weekl…

@EnergyCynic @WAR527 oh and they have to sell some stuff too to make it all work. just my opinion.

@EnergyCynic @WAR527 The land buy... not the merger. But in the end the land buy is probably correct by using todays large cashflows maintaining 70% CF back to shareholders and having extra to build out more volume in a normalized oil environment. Kind of solidified them as a good hold 3-5 years.

Could the $DVN/ $CTRA deal and subsequent acreage buy have been the short term market top?

@trend_bullish @HFI_Research pish posh - Cramer is known for his deep analysis and thoughtful commentary.

@HFI_Research I’ll say bottom at 75$… then we will hang between 75 and 95$ for the rest of tge year

Cramer calling for crude below $70 in a month.

I feel like the energy markets are in the subprime is contained phase of story.

@Mark4XX @grok find the full video

INTERESTING CLAIM: AMERICA ACTIVELY WANTS OIL PRICES TO EXPLODE TO UNBALANCE BRICS Geopolitical analyst Alexander Mercouris has uncovered a disturbing strategy allegedly brewing inside the United States. His sources point to heavy manipulation of financial markets designed to suppress oil prices artificially low, even while reserves are critically depleted. The prediction is clear. A sudden dramatic increase in oil prices is coming, yet some influential figures in Washington are actively encouraging this outcome. THE SHOCKING STRATEGY ➡️ Geopolitical analyst Alexander Mercouris claims the US government is “heavily manipulating the financial markets” to keep oil prices artificially low despite “critically depleted” reserves. ➡️ His source predicts a “sudden dramatic increase in oil prices” that could hit at any moment. ➡️ Some people in the United States not only are not concerned about this but actively want it to happen. ➡️ The reason is that it would “put extreme pressure on China” and “severely unbalance the BRICS.” THE RECESSION ACCEPTANCE ➡️ Mercouris adds that the alleged architects of this plan are “prepared to accept a global recession.” ➡️ Their aim is to “protect the dollar system” above everything else. ➡️ They also want to derail BRICS financial integration before it gains real strength. THE BOTTOM LINE Powerful interests in the United States may be willing to plunge the world into economic turmoil just to keep the dollar supreme and stop the BRICS from building a rival financial order. We shall soon see whether these forces will continue to pursue their plans even after the MOU, or whether it was all just a fabrication. #MercourisReveal #OilPriceShock #DollarDefense #BRICSUnderFire #GlobalRecessionRisk #ChinaTargeted #FinancialManipulation

I have been using 4.8 all day today... is fable better? yes - but it isn't the "model" - it is the redundancy layering uncapped agents and massive context ingestion that killed. At 1 point I had 200 agents working through a master orchestrator cross validating every little thing fable did. Ant decided to unlock that over opus. That is what people think is special automatic agentic spawning.

Hot take: a lot of people wouldn’t be able to tell the difference if they were randomly routed between gpt-5.5, opus-4.8, or fable-5 for their day to day work

depends on your needs right now. Current frontier models are heavily focused. They are not general by any stretch. I code mostly with gpt and claude, front end with claude for sure, but GPT is very good at clean code backend output. Gemini is very good at research due to googles piping. Grok can give you interesting x data depending on what you are looking for. Some open source models are doing really well for certain tasks as well. The gap is actually pretty small these days when looking at a overalls. Perplexity is pretty good choosing the right model for a task if you don't feel like testing them all.

@MaverickDots @Conor_D_Dart @AgenticPM Any idea what models they would be?

If the GPT 5.6 rumors are accurate, Anthropic may have a serious problem. • Better agentic coding than Fable • 3× lower cost • Fewer refusals • Stronger overall developer appeal I wonder if they will release GPT 5.6 sooner