01
Guess my election-interfering space lasers didn’t work this time 🤷♂️
Le PDG de Tesla et SpaceX rejette les théories du complot sur son influence suite aux derniers résultats des élections américaines.
Tendances sur X · Archives
mercredi 4 novembre 2026
Top 10 tweets · La sélection conservée telle qu’elle a été publiée.
- 02
People should always be in control of the future. OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”. People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
Le PDG d'OpenAI fixe une limite philosophique : l'action humaine doit rester la contrainte ultime dans le développement de l'IA.
- 03
Friday was my last day at Anthropic. Two years ago, when Claude Code was a small internal tool, I asked it to build an RL env. I steered it for tens of hours but never manually wrote a line of code. It worked, I was hooked, and I joined the team. When we first launched, most people approved every permission prompt by hand and mostly used Claude to answer questions and write tests. Today, the default is trusting Claude's judgment to auto-approve. Claude does large migrations and builds new products. Helping build Claude Code into an industry-defining product has been the journey of a lifetime. Thank you to the incredible team behind Claude Code, Cowork, and Tag. These products are the sum of thousands of great ideas from you. I'll miss working with you all. Thank you to our users for embracing a new way of working. Your constant feedback shaped so much of what we shipped and your stories made the work so meaningful. Excited to see what you all make next!
Un ingénieur quittant Anthropic souligne comment Claude Code est passé d'un outil interne à un agent autonome de création de produits.
- 04
ChatGPT and Codex are helping engineers build for the physical world. We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
OpenAI s'étend au matériel et à la fabrication, montrant comment ses modèles accélèrent l'ingénierie mécanique et électrique.
- 05
New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: https://www.anthropic.com/research/provable-inference Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://github.com/anthropics/FlashProve We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (http://rdalal.github.io) and Ben Mayer (@transmissions11).
Anthropic lance FlashProve, une méthode cryptographique pour vérifier les résultats de l'IA, ouvrant la voie à une IA d'entreprise sûre.
- 06
Morgan Stanley’s Adam Jonas in new @SpaceX note: “Just 4 months since our initiation, we’ve nearly doubled our 2027 revenue (to $175B) and EBIT estimates, while pulling forward FCF break-even by four years (to 2031) and cutting ~$280bn from our projected debt issuance. On fiscal year 2028, $SPCX trades cheaper than the S&P500. Interesting. Our financial forecasts have improved significantly: • Revenues: FY27 and FY28 revenues have been revised up 93% and 84% to $175B and $289B respectively. • Operating Income: Our FY28 EBIT forecast has been revised up 72% to $119B. • Free Cash Flow: At the time of our initiation, we had expected SPCX to be FCF negative until 2035. Today, we forecast the company to achieve positive FCF by 2031 - a 4 year pull-forward FCF break-even. • Expected Borrowing: At initiation, we anticipated SPCX would raise a combined $670B of external funds (IG borrowing) through the 2035 time horizon. Today, we forecast a combined $390B of borrowing through 2031. At our new 2028 earnings (adj.) of $109B and the current market cap of $1.91T ($140 a share), SPCX would trade at 17.6x P/E, cheaper than the S&P500 2028 forward P/E of 17.7x (MSe), while growing earnings at a ~150% CAGR from 2026-2030e.”
Morgan Stanley revoit à la hausse ses prévisions pour SpaceX, projetant une forte croissance des revenus et une rentabilité d'ici 2031.
- 07
Tesla's car registrations in Germany were up 290% YoY in Germany in October. The total number of Teslas registered between January and October 2026 rose 206% to 47,742 units compared with the same period last year. A couple of weeks ago, @Tesla's Giga Berlin factory built 7,000 Model Ys in one week for the first time. This year, Tesla expanded vehicle production at that factory from 5k vehicles per week to a goal of 7,500 per week due to higher demand, creating 2,000+ new jobs through early next year.
L'expansion de Tesla à Giga Berlin porte ses fruits avec un bond des immatriculations allemandes, prouvant la demande pour le Model Y.
- 08
Need a fresh start for your Windows 11 PC? With Cloud rebuild, you can download and reinstall Windows and the drivers your PC needs directly from Windows Update, even if Windows can’t start. No USB drive or installation media needed. Learn more: https://support.microsoft.com/windows/experience/backup-recovery/cloud-rebuild
Microsoft lance un outil de récupération cloud pour Windows 11, évitant le besoin de supports physiques pour restaurer un PC défaillant.
- 09
First Microsoft. Now Google. Microsoft opened a voluntary exit offer to about 60,000 people. That is over a quarter of the company. Google is now reportedly offering exit packages in Canada, France, Germany, Japan, and Switzerland. A few thoughts: 1. This is not a normal layoff. A voluntary offer lets a company shrink without the headline, and without the legal friction in places like France and Germany. Microsoft has also said it is not a replacement for layoffs, so I would expect both. 2. Follow the money. Microsoft expects around $175B in capex this year, mostly AI infrastructure. Data centers, chips and power are the costs going up fast. Headcount is the one line that can go down fast. 3. Voluntary has a catch. The people who take the package are often the ones with the most options. The company does not fully control who leaves. Microsoft's earlier retirement offer saw over 30% take it, so people do say yes. 4. The message to those who stay is clear: be all in on the AI strategy, or take the money. Google said close to this when it ran its earlier US programs. 5. Teams get smaller, expectations get higher, and AI tooling is expected to close the gap. Spend is going up while headcount goes down. The money is moving to AI infrastructure and to people who can ship with it. If you are an engineer, build things outside your employer's walls too. Big Tech used to compete on who hired the fastest. Now it competes on who can do the most with the fewest people.
Les géants de la tech utiliseraient des départs volontaires pour réduire discrètement leurs effectifs et compenser les coûts de l'IA.
- 10
$SPCX - SPACEX, TESLA SLIDE DESPITE MORGAN STANLEY'S BULLISH OUTLOOK SpaceX fell 5% and Tesla dropped 5.6% as Democratic election gains raised concerns over regulatory scrutiny and Elon Musk's political exposure. Meanwhile, Morgan Stanley raised its SpaceX price target to $300, implying approximately 114% upside from $140. 🔸 2027 revenue forecast: $175 billion 🔸 2028 revenue forecast: $289 billion 🔸 2028 operating income: $119 billion 🔸 Free-cash-flow breakeven: 2031, versus 2035 previously The bank sees substantial undervalued potential in SpaceX's AI operations.
Les marchés pèsent les prévisions optimistes de Morgan Stanley pour SpaceX face aux craintes réglementaires post-élections américaines.
- 01
Guess my election-interfering space lasers didn’t work this time 🤷♂️
Le PDG de Tesla et SpaceX rejette les théories du complot sur son influence suite aux derniers résultats des élections américaines.
- 03
Friday was my last day at Anthropic. Two years ago, when Claude Code was a small internal tool, I asked it to build an RL env. I steered it for tens of hours but never manually wrote a line of code. It worked, I was hooked, and I joined the team. When we first launched, most people approved every permission prompt by hand and mostly used Claude to answer questions and write tests. Today, the default is trusting Claude's judgment to auto-approve. Claude does large migrations and builds new products. Helping build Claude Code into an industry-defining product has been the journey of a lifetime. Thank you to the incredible team behind Claude Code, Cowork, and Tag. These products are the sum of thousands of great ideas from you. I'll miss working with you all. Thank you to our users for embracing a new way of working. Your constant feedback shaped so much of what we shipped and your stories made the work so meaningful. Excited to see what you all make next!
Un ingénieur quittant Anthropic souligne comment Claude Code est passé d'un outil interne à un agent autonome de création de produits.
- 05
New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: https://www.anthropic.com/research/provable-inference Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://github.com/anthropics/FlashProve We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (http://rdalal.github.io) and Ben Mayer (@transmissions11).
Anthropic lance FlashProve, une méthode cryptographique pour vérifier les résultats de l'IA, ouvrant la voie à une IA d'entreprise sûre.
- 07
Tesla's car registrations in Germany were up 290% YoY in Germany in October. The total number of Teslas registered between January and October 2026 rose 206% to 47,742 units compared with the same period last year. A couple of weeks ago, @Tesla's Giga Berlin factory built 7,000 Model Ys in one week for the first time. This year, Tesla expanded vehicle production at that factory from 5k vehicles per week to a goal of 7,500 per week due to higher demand, creating 2,000+ new jobs through early next year.
L'expansion de Tesla à Giga Berlin porte ses fruits avec un bond des immatriculations allemandes, prouvant la demande pour le Model Y.
- 09
First Microsoft. Now Google. Microsoft opened a voluntary exit offer to about 60,000 people. That is over a quarter of the company. Google is now reportedly offering exit packages in Canada, France, Germany, Japan, and Switzerland. A few thoughts: 1. This is not a normal layoff. A voluntary offer lets a company shrink without the headline, and without the legal friction in places like France and Germany. Microsoft has also said it is not a replacement for layoffs, so I would expect both. 2. Follow the money. Microsoft expects around $175B in capex this year, mostly AI infrastructure. Data centers, chips and power are the costs going up fast. Headcount is the one line that can go down fast. 3. Voluntary has a catch. The people who take the package are often the ones with the most options. The company does not fully control who leaves. Microsoft's earlier retirement offer saw over 30% take it, so people do say yes. 4. The message to those who stay is clear: be all in on the AI strategy, or take the money. Google said close to this when it ran its earlier US programs. 5. Teams get smaller, expectations get higher, and AI tooling is expected to close the gap. Spend is going up while headcount goes down. The money is moving to AI infrastructure and to people who can ship with it. If you are an engineer, build things outside your employer's walls too. Big Tech used to compete on who hired the fastest. Now it competes on who can do the most with the fewest people.
Les géants de la tech utiliseraient des départs volontaires pour réduire discrètement leurs effectifs et compenser les coûts de l'IA.
- 02
People should always be in control of the future. OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”. People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
Le PDG d'OpenAI fixe une limite philosophique : l'action humaine doit rester la contrainte ultime dans le développement de l'IA.
- 04
ChatGPT and Codex are helping engineers build for the physical world. We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
OpenAI s'étend au matériel et à la fabrication, montrant comment ses modèles accélèrent l'ingénierie mécanique et électrique.
- 06
Morgan Stanley’s Adam Jonas in new @SpaceX note: “Just 4 months since our initiation, we’ve nearly doubled our 2027 revenue (to $175B) and EBIT estimates, while pulling forward FCF break-even by four years (to 2031) and cutting ~$280bn from our projected debt issuance. On fiscal year 2028, $SPCX trades cheaper than the S&P500. Interesting. Our financial forecasts have improved significantly: • Revenues: FY27 and FY28 revenues have been revised up 93% and 84% to $175B and $289B respectively. • Operating Income: Our FY28 EBIT forecast has been revised up 72% to $119B. • Free Cash Flow: At the time of our initiation, we had expected SPCX to be FCF negative until 2035. Today, we forecast the company to achieve positive FCF by 2031 - a 4 year pull-forward FCF break-even. • Expected Borrowing: At initiation, we anticipated SPCX would raise a combined $670B of external funds (IG borrowing) through the 2035 time horizon. Today, we forecast a combined $390B of borrowing through 2031. At our new 2028 earnings (adj.) of $109B and the current market cap of $1.91T ($140 a share), SPCX would trade at 17.6x P/E, cheaper than the S&P500 2028 forward P/E of 17.7x (MSe), while growing earnings at a ~150% CAGR from 2026-2030e.”
Morgan Stanley revoit à la hausse ses prévisions pour SpaceX, projetant une forte croissance des revenus et une rentabilité d'ici 2031.
- 08
Need a fresh start for your Windows 11 PC? With Cloud rebuild, you can download and reinstall Windows and the drivers your PC needs directly from Windows Update, even if Windows can’t start. No USB drive or installation media needed. Learn more: https://support.microsoft.com/windows/experience/backup-recovery/cloud-rebuild
Microsoft lance un outil de récupération cloud pour Windows 11, évitant le besoin de supports physiques pour restaurer un PC défaillant.
- 10
$SPCX - SPACEX, TESLA SLIDE DESPITE MORGAN STANLEY'S BULLISH OUTLOOK SpaceX fell 5% and Tesla dropped 5.6% as Democratic election gains raised concerns over regulatory scrutiny and Elon Musk's political exposure. Meanwhile, Morgan Stanley raised its SpaceX price target to $300, implying approximately 114% upside from $140. 🔸 2027 revenue forecast: $175 billion 🔸 2028 revenue forecast: $289 billion 🔸 2028 operating income: $119 billion 🔸 Free-cash-flow breakeven: 2031, versus 2035 previously The bank sees substantial undervalued potential in SpaceX's AI operations.
Les marchés pèsent les prévisions optimistes de Morgan Stanley pour SpaceX face aux craintes réglementaires post-élections américaines.
- 01
Guess my election-interfering space lasers didn’t work this time 🤷♂️
Le PDG de Tesla et SpaceX rejette les théories du complot sur son influence suite aux derniers résultats des élections américaines.
- 04
ChatGPT and Codex are helping engineers build for the physical world. We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
OpenAI s'étend au matériel et à la fabrication, montrant comment ses modèles accélèrent l'ingénierie mécanique et électrique.
- 07
Tesla's car registrations in Germany were up 290% YoY in Germany in October. The total number of Teslas registered between January and October 2026 rose 206% to 47,742 units compared with the same period last year. A couple of weeks ago, @Tesla's Giga Berlin factory built 7,000 Model Ys in one week for the first time. This year, Tesla expanded vehicle production at that factory from 5k vehicles per week to a goal of 7,500 per week due to higher demand, creating 2,000+ new jobs through early next year.
L'expansion de Tesla à Giga Berlin porte ses fruits avec un bond des immatriculations allemandes, prouvant la demande pour le Model Y.
- 10
$SPCX - SPACEX, TESLA SLIDE DESPITE MORGAN STANLEY'S BULLISH OUTLOOK SpaceX fell 5% and Tesla dropped 5.6% as Democratic election gains raised concerns over regulatory scrutiny and Elon Musk's political exposure. Meanwhile, Morgan Stanley raised its SpaceX price target to $300, implying approximately 114% upside from $140. 🔸 2027 revenue forecast: $175 billion 🔸 2028 revenue forecast: $289 billion 🔸 2028 operating income: $119 billion 🔸 Free-cash-flow breakeven: 2031, versus 2035 previously The bank sees substantial undervalued potential in SpaceX's AI operations.
Les marchés pèsent les prévisions optimistes de Morgan Stanley pour SpaceX face aux craintes réglementaires post-élections américaines.
- 02
People should always be in control of the future. OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”. People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
Le PDG d'OpenAI fixe une limite philosophique : l'action humaine doit rester la contrainte ultime dans le développement de l'IA.
- 05
New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: https://www.anthropic.com/research/provable-inference Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://github.com/anthropics/FlashProve We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (http://rdalal.github.io) and Ben Mayer (@transmissions11).
Anthropic lance FlashProve, une méthode cryptographique pour vérifier les résultats de l'IA, ouvrant la voie à une IA d'entreprise sûre.
- 08
Need a fresh start for your Windows 11 PC? With Cloud rebuild, you can download and reinstall Windows and the drivers your PC needs directly from Windows Update, even if Windows can’t start. No USB drive or installation media needed. Learn more: https://support.microsoft.com/windows/experience/backup-recovery/cloud-rebuild
Microsoft lance un outil de récupération cloud pour Windows 11, évitant le besoin de supports physiques pour restaurer un PC défaillant.
- 03
Friday was my last day at Anthropic. Two years ago, when Claude Code was a small internal tool, I asked it to build an RL env. I steered it for tens of hours but never manually wrote a line of code. It worked, I was hooked, and I joined the team. When we first launched, most people approved every permission prompt by hand and mostly used Claude to answer questions and write tests. Today, the default is trusting Claude's judgment to auto-approve. Claude does large migrations and builds new products. Helping build Claude Code into an industry-defining product has been the journey of a lifetime. Thank you to the incredible team behind Claude Code, Cowork, and Tag. These products are the sum of thousands of great ideas from you. I'll miss working with you all. Thank you to our users for embracing a new way of working. Your constant feedback shaped so much of what we shipped and your stories made the work so meaningful. Excited to see what you all make next!
Un ingénieur quittant Anthropic souligne comment Claude Code est passé d'un outil interne à un agent autonome de création de produits.
- 06
Morgan Stanley’s Adam Jonas in new @SpaceX note: “Just 4 months since our initiation, we’ve nearly doubled our 2027 revenue (to $175B) and EBIT estimates, while pulling forward FCF break-even by four years (to 2031) and cutting ~$280bn from our projected debt issuance. On fiscal year 2028, $SPCX trades cheaper than the S&P500. Interesting. Our financial forecasts have improved significantly: • Revenues: FY27 and FY28 revenues have been revised up 93% and 84% to $175B and $289B respectively. • Operating Income: Our FY28 EBIT forecast has been revised up 72% to $119B. • Free Cash Flow: At the time of our initiation, we had expected SPCX to be FCF negative until 2035. Today, we forecast the company to achieve positive FCF by 2031 - a 4 year pull-forward FCF break-even. • Expected Borrowing: At initiation, we anticipated SPCX would raise a combined $670B of external funds (IG borrowing) through the 2035 time horizon. Today, we forecast a combined $390B of borrowing through 2031. At our new 2028 earnings (adj.) of $109B and the current market cap of $1.91T ($140 a share), SPCX would trade at 17.6x P/E, cheaper than the S&P500 2028 forward P/E of 17.7x (MSe), while growing earnings at a ~150% CAGR from 2026-2030e.”
Morgan Stanley revoit à la hausse ses prévisions pour SpaceX, projetant une forte croissance des revenus et une rentabilité d'ici 2031.
- 09
First Microsoft. Now Google. Microsoft opened a voluntary exit offer to about 60,000 people. That is over a quarter of the company. Google is now reportedly offering exit packages in Canada, France, Germany, Japan, and Switzerland. A few thoughts: 1. This is not a normal layoff. A voluntary offer lets a company shrink without the headline, and without the legal friction in places like France and Germany. Microsoft has also said it is not a replacement for layoffs, so I would expect both. 2. Follow the money. Microsoft expects around $175B in capex this year, mostly AI infrastructure. Data centers, chips and power are the costs going up fast. Headcount is the one line that can go down fast. 3. Voluntary has a catch. The people who take the package are often the ones with the most options. The company does not fully control who leaves. Microsoft's earlier retirement offer saw over 30% take it, so people do say yes. 4. The message to those who stay is clear: be all in on the AI strategy, or take the money. Google said close to this when it ran its earlier US programs. 5. Teams get smaller, expectations get higher, and AI tooling is expected to close the gap. Spend is going up while headcount goes down. The money is moving to AI infrastructure and to people who can ship with it. If you are an engineer, build things outside your employer's walls too. Big Tech used to compete on who hired the fastest. Now it competes on who can do the most with the fewest people.
Les géants de la tech utiliseraient des départs volontaires pour réduire discrètement leurs effectifs et compenser les coûts de l'IA.
Classés par engagement. Les publications appartiennent à leurs auteurs sur X. Chiffres relevés lors de la préparation de cette sélection.