| | Good morning. Scale AI very quietly trimmed its remote headcount by about 1,000 contractors. | The Wall Street Journal also reported that OpenAI is in talks to raise more money, now at a valuation in excess of $100 billion. | And yes, Nvidia reported earnings last night. And the stock is in the red. | — Ian Krietzberg, Editor-in-Chief, The Deep View | In today’s newsletter: | |
| |
| AI for Good: (Tracking) Twisters | | Source: Unsplash |
| Scientists don’t yet know how, why (and when) tornadoes form, something that makes them extraordinarily difficult to forecast — in examining identical storm systems, “one will produce a tornado and one won't. We don't fully understand it,” according to James Kurdzo, a principal investigator on MIT’s TorNet project. | Artificial intelligence — specifically deep learning — might lay the foundation for an answer. | The details: Earlier this year, MIT researchers open-sourced a dataset called TorNet, in addition to several AI models trained on that dataset. (TorNet contains hundreds of thousands of radar images of storm systems and tornadoes). | The baseline deep learning model the researchers developed performed on par or better than existing tornado-detection systems: it “classified 50% of weaker EF-1 tornadoes and over 85% of tornadoes rated EF-2 or higher, which make up the most devastating and costly occurrences of these storms.” (The current rate of false alarms for tornado warnings is above 70%). The key to its adoption is explainability; researchers need to know how and why the model came to certain conclusions.
| Why it matters: Kurdzo hopes the advancement will help forecasters make better predictions, something that could return trust in tornado warnings and increase evacuation time. He also hopes the system can help unravel the science behind why tornadoes form. |
| |
| | The Ibuprofen Alternative You Need to Know | | Stomach pain, dizziness, heart concerns, kidney problems… | These are just some of the scary side effects that can be found on the back of a bottle of Ibuprofen. | And to top it all off, Ibuprofen doesn’t even address the main cause of your problem… It’s only temporary pain relief. | You don't need to just alleviate your pain–you need to fight INFLAMMATION. That’s why researchers speculate this one natural food could outperform Ibuprofen… | And it doesn’t have the dangerous side effects mentioned above either! | THIS will erase your pain (forget Ibuprofen) |
| |
| Advocates accuse Elon Musk’s xAI of adding to Memphis’ smog problems | | Source: Unsplash |
| The Southern Environmental Law Center (SELC), in a letter sent this week to the EPA and local health department, accused Elon Musk’s Memphis-based xAI data center of adding to Memphis’ air pollution problems by operating gas-powered turbines without a permit. | The details: The letter says that xAI “has installed at least 18 gas combustion turbines over the last several months (with more potentially on the way).” | | The context: xAI — which has achieved a $24 billion valuation — opened the center in June and started training a new model on 100,000 Nvidia GPUs there in July. | | The advocates added at the time that “the environmental impacts of providing power to the xAI facility, including air pollution, water and environmental justice impacts, have not yet been studied as required by the National Environmental Policy Act.” | Further, the letter said that it would require $760,000 for the local power company, MLGW, to increase power to service xAI, plus $24 million to build a new substation. According to the letter, the power company plans to allow xAI a “marginal allowance” to recoup those costs, meaning local ratepayers would be subsidizing xAI’s power needs. | MLGW told CNBC that it started supplying xAI with 50 MW of power at the beginning of August. |
| |
| | AI Panel: How AI automation is transforming industries | | Join us for a panel discussion with leading experts from Glean, Braintrust, and Monterey AI to learn about AI's challenges, opportunities and its transformative potential for the future of work. |
| |
| | | | | Goldman issues warning on how quickly market confidence has recovered from an August stock slump (CNBC). TikTok must face lawsuit over 10-year-old girl's death, US court rules (Reuters). Scale AI lays off workers via email with no warning (Inc). The rise and fall of OpenSea (The Verge).
| Growth Forum: Join 20,000+ leaders getting smarter B2B sales. Plus every week, we share firsthand experiments, expert interviews, and actionable advice designed to help you sell more. | If you want to get in front of an audience of 200,000+ developers, business leaders and tech enthusiasts, get in touch with us here. | | | | | | |
| |
| California passes AI protection bill | | Source: Created with AI by The Deep View |
| A bill that would protect performers from unauthorized AI replicas this week passed California’s state Senate. It will soon head toward the governor’s desk. | The details: The bill, AB 2602, requires employers to explicitly bargain for the rights to use a performer’s AI replica in the initial contractual agreement. The employer’s intentions with the replica must be specifically outlined – and agreed to — or else the contract is unenforceable. | The bill passed the Senate 37-1; it passed the Assembly 62-0 in May. “Voice and likeness rights, in an age of digital replication, must have strong guardrails around licensing to protect from abuse, this bill provides those guardrails,” SAG-AFTRA executive director Duncan Crabtree-Ireland told Variety. SAG-AFTRA is currently striking video game companies over AI protections.
| The context: | GenAI has made the idea of protecting digital likenesses paramount for artists across mediums. Earlier this year, Tennessee passed the ELVIS Act, legislation that protects an artist’s voice from unauthorized synthetic replication.
| At the same time, the infamous SB 1047 passed the state Assembly and now moves back to the Senate before making its way to the governor’s desk. |
| |
| Nvidia stock takes a dive despite earnings beat | | Source: Nvidia |
| Chipmaker Nvidia reported earnings Wednesday night. Wedbush analyst Dan Ives called it the “most important earnings report” of the year. In the leadup to the report, the stock closed the day down about 2%, then went on to fall another 7% in after-hours trading. | Here’s how Nvidia performed this quarter: Nvidia posted revenue of $30.0 billion, above expectations of $28.7 billion (that’s an increase of 122% from a year ago). The semiconductor reported adjusted earnings of 68 cents per share, above expectations of 64 cents. | Revenue from Nvidia’s data center business (which encompasses its AI chips) spiked 154% year-over-year to $26.3 billion, also topping analyst expectations. Nvidia said it expects $32.5 billion in revenue for the current quarter, above expectations of $31.7 billion.
| This marks Nvidia’s narrowest revenue beat in six quarters, something that could well have contributed to the market’s immediate negative reaction. | “Hopper demand remains strong, and the anticipation for Blackwell is incredible,” CEO Jensen Huang said in a statement. | In response to analyst questions about whether the Big Tech companies buying Nvidia chips will see a return on their enormous capital expenditures, Huang said: “It’s the best ROI on a computing infrastructure investment you can make today.” | What the analysts are saying: Global X’s ETFs research analyst Ido Caspi said that, despite the deceleration in revenue growth, “122% growth at Nvidia’s scale is still significant and speaks to a robust AI investment environment. Big Tech's capital expenditure remains exceptionally high given how early we are in the generative AI investment cycle.” | Deepwater’s Gene Munster said that the general commentary on the report and the call is that “things were good, they weren’t great, and you gotta have the stock be great for it to continue to work.” But he added that, from his perspective, “nothing changed.” He said the debate should include a question: “to what degree do you think AI is going to be transformative?” According to Munster, if you think AI will be as transformative as the internet, then buying Nvidia doesn’t make sense; if you think it’ll be more transformative, “then you see a case that this growth will continue well into 2026.”
| Ives said that the negative stock reaction is “just noise” given Nvidia’s numbers. The AI Revolution, he said, has begun. | The broader impact: Even with the dip, Nvidia is still among the three largest companies in the world. The impact of this stock is therefore pretty broadly felt, both visibly in the markets and less tangibly as a bellwether for the reality of the AI boom. | | | | Earlier this week, we talked about Nvidia as a beacon for the AI investment. What this beacon is indicating is simple: expectations are way too high, and patience is wearing thin. | On paper, Nvidia had a great quarter. Demand looks strong, revenue, yet again, set a record … yet the stock shed hundreds of billions in value. Whether it gains it all back tomorrow remains to be seen, and doesn’t really matter. Volatility and investor sentiment both are clear. And concern about ROI on the AI investment, not for Nvidia, but for Nvidia’s customers, remains paramount. Nvidia is the canary in the coal mine; if, for instance, Microsoft’s AI isn’t paying off, it’ll stop buying GPUs and Nvidia’s data center revenues will be the first thing to fall.
| That hasn’t happened, but it seems to be on investors’ minds. | Munster nailed it — the Nvidia debate absolutely depends on how transformative you think AI will be. I’ll answer his question with a skeptical one of my own: what happens to the pick-and-shovel companies when everyone else realizes there is no gold? | | | Which image is real? | | | | | 🤔 Your thought process: | Selected Image 2 (Left): | | Selected Image 1 (Right): | “I don't believe you.” |
| |
| 💭 A poll before you go | Thanks for reading today’s edition of The Deep View! | We’ll see you in the next one. | Here’s your view on OpenAI’s Strawberry: | More than a third of you said it sounds powerful now, but in reality, it won’t be. 20% said it’s straight hype and 15% said it’s another path on the road to AGI. Another 15% said, hype or not, OpenAI shouldn’t be allowed to release it. | Sounds powerful now, but in reality, it won’t be: | | What do you think about Nvidia's tumble? | |
|
|
|