← Tech Brew Ride Home
Tech Brew Ride Home
My LLM Cutoff Is 1930
00:21:11
Apple beat Q2 estimates and forecast strong Q3 growth as the MacBook Neo and Mac Mini sell out. The Senate unanimously banned prediction market trading for senators, Intel closed its best month ever at +114%, and Musk admitted xAI "partly" distilled OpenAI models.
Loading summary
Transcript4 lines
- [00:00]
A
Study and play come together on a Windows 11 PC and for a limited time, college students get the best of both worlds. Get the unreal college deal Everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass ultimate with a custom color Xbox wireless controller. Learn more@windows.com studentoffer while supplies last ends June 30th terms at aka mscollegepc.
- [00:35]
B
Welcome to the Tech Brew. Write Home for Friday, May 1, 2026 I'm Brian McCullough. Today, Apple beat Q2 estimates and forecast strong Q3 growth as the MacBook Neo and Mac Mini sell out. The Senate unanimously banned prediction market trading for senators. Intel closed its best month ever at greater than 114% up, Elon admitted XAI partly distilled OpenAI models and of course the weekend long read suggestions. Here's what you missed today in the world of TEC Apple earnings. They did good. Yada yada yada. Apple's CFO said that the iPhone 17 family is now the most popular lineup in our history and we believe we gained market share during the quarter. And Tim Cook said iPhone sales, which slightly missed Q2 estimates, were held back by supply chain constraints for advanced chips as quote, demand was off the charts. But Apple is up 5% as I typed this this morning because Apple forecast Q3 revenue above estimates with sales expected to rise between 14 and 17%. They did also say though, that memory expenses will climb significantly higher in Q3, quoting Bloomberg. The outlook bodes well for incoming Chief Executive Officer John Ternus, who takes the reins from Tim Cook on September 1st. Ternus appeared briefly on the conference call saying he would maintain Cook's quote, thoughtfulness, deliberateness and discipline as CEO. We have an incredible roadmap ahead, turner said. This is the most exciting time in my 25 year career at Apple to be building products and services. Cook, who has run the Cupertino, California based company for 15 years, will remain as Executive chairman. Apple has benefited from a series of new products launched in March, including the MacBook Neo, iPhone 17e, updated iPad Air models and a fresh MacBook Pro. The $599 Neo, Apple's first major push into low cost laptops, has been particularly popular and remains sold out at retailers. Total sales gained 17% to $111.2 billion during the fiscal second quarter, which ended March 28. Analysts had anticipated $109.7 billion on average. Apple itself had projected sales growth of 13 to 16%. Apple signaled that it's mostly coping with shortages and memory costs, but things will worsen over time. A memory crunch has rippled through the tech industry, forcing companies to boost prices and reduce output. On the call, Cook said memory expenses would climb significantly higher this quarter compared with more subdued increases in the March period. In the fourth quarter and beyond, he said there would be an increasing impact on the business. He declined to disclose if Apple would raise prices. Apple's main supply constraints involve processors rather than memory chips. The impact has primarily hit the Mac Mini and Mac Studio, two desktop Macs that have become popular for running artificial intelligence models. Cook said the company under called demand on those Macs. It also underestimated the Appetite for the MacBook Neo, which also remains hard to find. Apple's online store quotes several week waits for all three products and sellouts for some configurations. We're not at the point where we're saying this is going to end anytime soon, cook said, adding that the constraints will likely last several months. He characterized the iPhone shortages as being less significant than the ones affecting the Mac. As part of its second quarter earnings report, Apple said it would buy back as much as $100 billion in shares and boosted its dividend. The company also said it would be winding down its net CA strategy as it begins to look at cash and debt independently. Earnings rose $2.01 a share, beating the projection of $1.96. The China market, where Apple has struggled in recent years, was a highlight. Last quarter, revenue there soared 28% to $20.5 billion. Wall street had projected only 18.9 billion. Amazon has debuted Join the Chat, an AI powered feature that lets users ask questions about products and get conversational audio responses generated in real time. Quoting TechCrunch, Amazon launched a new AI powered feature on Tuesday that allows users to ask questions about products and receive conversational audio responses. The responses are delivered by what the company calls AI powered shopping experts, which present information in a natural discussion style format. The new Join the Chat feature aims to save customers time by providing key product details without requiring them to scroll through lengthy descriptions or reviews. The AI pulls together insights about product features, customer feedback and other relevant information. For example, shoppers can ask questions like whether a coffee maker is suited for beginners or whether a sweater feels itchy based on customer reviews. Rather than giving generic answers, Amazon says the AI builds on previous responses to provide more relevant and helpful information, while also making sure not to repeat anything. This is meant to be a similar experience to speaking with a knowledgeable employee at a IRL Store customers can ask questions and actually steer where the conversation goes. Every question they ask influences what comes next, making the experience a conversation customers can join and customize. The company writes on a blog post. Join the chat as part of a broader experience called Hear the Highlights, which offers short audio summaries on millions of product pages within the Amazon Shopping app. That feature began testing last May and is currently available in the U.S. however, only select products have audio summaries. To use the feature, customers open a product page in the app and tap the Hear the Highlights button located below the product image. From there, they can listen to a brief overview or tap the Join the Chat icon to ask specific questions via text or voice. The audio can continue playing even as users browse. The new capability builds on Amazon's growing lineup of AI driven shopping tools. These include Rufus, its generative AI assistant that helps customers research products and compare options Interests, which continuously tracks and services new items aligned with a shopper's preferences and Help Me Decide, which suggests products based on a person's searches, browsing and shopping history. End quote. The United States Senate has unanimously passed a rule barring senators from trading on prediction markets like CalSHI and Polymarket amid rising concerns over insider trading. Quoting CNBC, the move came amid rising concern about insider trading on prediction market platforms like Kalshi and polymarket, and about event contracts that can involve death or violence. On April 22, Kalshee said it had suspended and fined one U.S. senate candidate and two candidates for the House of Representatives for political insider trading on their own campaigns. On April 23, a U.S. army Special Forces soldier, Master Sergeant Gannon Ken Van Dyke, was arrested on an indictment accusing him of classified information to make bets on polymarket related to the American military mission that captured Venezuelan leader Nicolas Maduro. Van Dyke, who was involved in that mission, won nearly $410,000 on those wagers, according to the Department of Justice. Earlier on Thursday, a group of Democratic members of Congress called on the Commodity Futures Trading Commission to issue a rule that prevents insider trading and corruption in the market and prohibits event contracts on the outcome of elections, war and military actions in the US or abroad, sports and government actions without a valid economic hedging interest. Kalshi and polymarket both praise the Senate's action. I applaud the Senate for passing this resolution to ban senators and their offices from trading on prediction markets, kalshi CEO Trek Mansour wrote in a post on X. Kalshi already proactively blocks members of Congress and enforces against insider trading. This is a great step to increase trust in our markets by making it an industry standard, mansoor said. Now let's pass this in the house. Polymarket in its own post on X said, we're in full support of this. Our rulebook and terms of service already prohibit such conduct, but codifying this into law as a step forward for the industry. Happy to help move this forward, however, we can End quote. One little Musk v. Altman nugget from the trial for you today. Quoting Wired While testifying on Thursday in federal court, Elon Musk seemed to indicate that his AI lab may have used OpenAI's models to train Xai's own. He touched upon the topic while sitting on the witness stand answering cross examination questions from an OpenAI attorney amid his ongoing legal battle against the ChatGPT maker. This is the exchange best as Wired could Capture it. QUOTE OpenAI Lawyer William Sabbat do you know what distillation is? Musk it means to use one AI model to train another AI model. Sabbat Has XAI done that with OpenAI? Musk generally all the AI companies do that. Sabbat so that's a yes. Musk Partly Distillation is a technique where a smaller AI model is trained to mimic the behavior of a larger, more capable model, making it cheaper and faster to run while preserv its performance. OpenAI's lawyer William Savitt then asked whether OpenAI's technology had been used in any way to develop XAI. Savit Has OpenAI technology been used in any way to develop Xai? Musk it is standard practice to use other AIs to validate your AI. OpenAI and Xai did not immediately respond to Wire's request for comment. OpenAI has been trying to prevent its competitors from distilling its AI models, in particular the Chinese AI lab Deep seq. In a February 2026 memo to a House committee, OpenAI wrote that it has taken steps to protect harden our models against distillation. In that memo, OpenAI said it was focused on ensuring a playing field in which China can't advance autocratic AI by appropriating and repackaging American innovation. End quote.
- [10:24]
C
The Wired newsroom is known for award winning reporting on how technology shapes our world. On WIRED's Uncanny Valley, we take that curiosity even further. Each week, journalists from Wired break down the biggest stories in tech while speaking directly with the people building challenging and reshaping the future. Is the AI boom sustainable? How do you protect your privacy in an age of constant surveillance? Uncanny Valley tackles the questions driving today's tech debates and lighting up your group chats. Listen to new episodes every Thursday wherever you get your podcasts.
- [10:56]
B
I have a long standing policy on this pod that if I tell you about the bad times when someone is on the way down, I owe it to you to follow up and let you know if maybe they've begun a road to recovery. Did you know Intel's stock jumped 114% in April? Quoting CNBC intel is on a winning streak unlike any since it became one of the first companies to go public on the Nasdaq nearly 55 years ago. The chipmaker stock soared 114% in April, closing out its best month on record. It's been an extended rally for intel, which has enjoyed two of its best days ever in the last seven months, including a 24% jump on April 24 following a blowout earnings report. The stock rose to a record that day for the first time since 2000 and then continued to rise. Intel is in the midst of a turnaround after years of delayed launches and disappointing yields that saw it fall far behind manufacturing leader Taiwan Semiconductor and chipmaker Nvidia in the race to power artificial intelligence. Wall street appears convinced the tide may be turning, with Intel's latest 18A chips showing real promise as they churn out of the company's new Arizona plant. At the same time, agentic AI is spinning up a major resurgen in demand for Intel's core product, the central processing unit. Bank of America predicts the CPU market could more than double by 2030, and Nvidia told CNBC in March that CPUs are becoming the bottleneck for AI. The CPU is reinserting itself as the indispensable foundation of the AI era, intel CEO Lip Bhutan said on the company's earnings call last week, adding that demand for ITS data center CPUs exceeds supply. Tan was tapped as CEO in March of 2025, three months after intel ousted Pat Gelsinger, who's year tenure was marred by turmoil. Intel's stock plummeted 60% in 2024, its worst year ever. Since then, it's almost quintupled, lifting Intel's market cap past $470 billion. While Intel's financials are showing signs of recovery, investors are getting way out ahead of fundamentals. Revenue in the latest quarter rose more than 7% after declines in five of the prior seven periods. But demand is materializing, driven by a scramble for compute by Intel's major hyperscaler customers Google, Microsoft and Amazon, as well as equipment manufacturers such as Dell, HP and Lenovo CPUs are cool again, and intel can't make enough, Moorinsight CEO Patrick Moorhead, who's been covering intel for 35 years, told CNBC in an interview. They're sold out, and they were able to raise prices. Intel's latest CPU for PCs, the Core Ultra Series 3, began selling in January, while its newest Xeon 6 Plus data center CPUs hit the market in March. The stock rally began months earlier after the US Government threw a lifeline to destroy struggling chipmaker in August by taking a 10% stake in the company and becoming its biggest shareholder. The Trump administration's $8.9 billion investment primarily comes from grants promised under the Chips act, signed by President Joe Biden in 2022. The government's stake in intel is now worth over $40 billion. Intel's real turning point began years ago when Gelsinger put renewed focus on the manufacturing side of the business, known as a foundry. Unlike fellow chip makers Advanced Micro Devices and Nvidia, which outsourced the complex and expensive manufacturing of their silicon, intel both designs and manufactures its own chips with the hopes of manufacturing for others as well. So far, intel remains the only major customer of its foundry, as longtime TSMC customers are hesitant to make the leap. Moorhead estimated that 75% of their valuation is about foundry and the promise of Foundry, which they have not yet delivered on. Tan has moved to unwind some of Gelsinger's other more aggressive efforts. Intel slashed 15% of its workforce in July and canceled Chip Fab projects in Germany and Poland. In Ohio, Intel's giant new Chip Fab is delayed until 2030 after initial plans had its starting production this year. Tan wrote in a memo about the layoffs that over the past several years the company invested too much too soon without adequate demand. In January, though, Tan began to change his tune, saying intel is going big time into its Next Generation Technology 14A. Tan said on last week's earnings call that multiple customers are actively evaluating the technology and it's being developed at a faster pace than 18a. Intel's only major outside commitment for foundry so far came from Elon Musk. Intel announced earlier this month that it would be joining Musk's terrafab chip complex in Austin, Texas to help design, fabricate and package ultra high performance chips at scale for SpaceX, XAI and Tesla. During Tesla's first quarter earnings call, Musk said Tesla plans to use Intel's forthcoming 14A process to produce chips at the facility, which is meant to make chips for use in Tesla's vehicles and robots and in yet to be constructed orbital Data centers for SpaceX, Moorhead said. Musk's announcement, even if vague, is what made Intel's stock absolutely pop. And the weekend Long Reads this weekend why would you train a language model on stuff that only happened before the year 1930? Well, I mean, what would happen if you did that? Because seriously, I'm curious as hell. And I'm not the only one quoting talking LM. For example, we can evaluate LM's ability to predict the future. Inspired by Calcifer Computing's work on temporal language models, we calculated the surprisingness of short descriptions of historical events to a 13B model trained on pre1931 text. We can see an increase after the knowledge cutoff, particularly pronounced in the 1950s and 1960s, followed by a plateau. We will continue to develop evals to measure with greater confidence how forecasting performance improves with model size and decays at longer horizons. Training larger vintage language models will allow us to uncover these scaling trends. Similarly, we can test LM's abilities to come up with new ideas by seeing if they can arrive at inventions or scientific discoveries we know would arise after their knowledge cutoffs, such as those pictured in Figure 2. As Demis Hassibas has asked, could a model trained up to 1911 independently discover general relativity, as Einstein did in 1915? End quote. Fascinating. No bonus episodes for you this weekend. Talk to you on Monday.