Artwork

Το περιεχόμενο παρέχεται από το owith. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον owith ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.
Player FM - Εφαρμογή podcast
Πηγαίνετε εκτός σύνδεσης με την εφαρμογή Player FM !

Tech Turmoil: From Orion Disappointment to Regulatory Uncertainty and AI Arms Race

3:13
 
Μοίρασέ το
 

Manage episode 449710189 series 3456961
Το περιεχόμενο παρέχεται από το owith. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον owith ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.
Good morning from OWITH.ai: the podcast that gives you only what's important to hear in AI and tech world.In March 2024, OpenAI CEO Sam Altman hinted at the upcoming GPT-5 model, but it fell short of expectations. The scaling laws that AI companies relied on for improvement are now breaking down, as seen with the underwhelming performance of the Orion model. Cognitive scientist Gary Marcus had previously warned about the limitations of deep learning, and his concerns are now being validated. OpenAI's new direction with the O1 model, focusing on real-time pondering and improved reasoning, represents a shift away from traditional GPT models. The success of this new paradigm remains uncertain, as it may require more time and resources to generate intelligent responses. In March 2025, the release of Orion leads to disappointment among users, shareholders, and critics, signaling the potential end of generative AI.Lina Khan's departure as chair of the Federal Trade Commission under the new Trump administration may not be universally celebrated in the tech industry. Trump's impact on tech regulation is uncertain, with potential changes in the crypto industry and state-level regulations playing a significant role. Despite speculation about increased M&A activity in the tech industry post-Khan, the actual impact of her antitrust lawsuits on VC portfolios is questioned. The battles over technology oversight are likely to occur at the state and municipal level over the next four years. Recent venture deals include companies like Parker and UnifyApps raising significant funding rounds. The future of tech regulation under Trump remains uncertain, with potential changes in the crypto industry and state-level regulations playing a significant role in shaping the landscape.The US has imposed export restrictions on advanced chips to China, leading Taiwan's TSMC to suspend chip shipments. Swiggy's IPO in India is oversubscribed, despite the company not yet being profitable. Chegg is facing a decline as students turn to ChatGPT for homework help. Visa and Tencent are launching a palm-based authentication system for payments in Singapore. Apple is expected to focus on wearables for future growth, moving away from the iPhone. Amazon is changing its physical stores, and an AI painting sold for a high price at auction.The email is a summary of the 2024 Fortune Global Forum in New York City, featuring speakers such as Mike Pompeo, Leon Panetta, and Boris Johnson. Amazon CEO Andy Jassy plans to reduce bureaucracy by increasing the ratio of employees to managers. Market research suggests bullish market performance could extend through the 2030s with Republican control of the government.The text discusses the increasing alignment between Silicon Valley and the Pentagon, leading to the militarization of AI and the development of dystopian applications such as automated weapons. The U.S. is leading in drone technology, with big tech companies partnering with defense agencies to compete in the global AI arms race.The text discusses the possibility that the core hypothesis driving generative AI, particularly in terms of scaling, may be incorrect. The author argues that deep learning may not improve with scale for tasks such as reasoning and handling edge cases in open-ended environments. Recent evidence suggests that pure scaling may have reached a point of diminishing returns.

Thanks for listening! Follow us on Twitter, Instagram and Linkedin

  continue reading

82 επεισόδια

Artwork
iconΜοίρασέ το
 
Manage episode 449710189 series 3456961
Το περιεχόμενο παρέχεται από το owith. Όλο το περιεχόμενο podcast, συμπεριλαμβανομένων των επεισοδίων, των γραφικών και των περιγραφών podcast, μεταφορτώνεται και παρέχεται απευθείας από τον owith ή τον συνεργάτη της πλατφόρμας podcast. Εάν πιστεύετε ότι κάποιος χρησιμοποιεί το έργο σας που προστατεύεται από πνευματικά δικαιώματα χωρίς την άδειά σας, μπορείτε να ακολουθήσετε τη διαδικασία που περιγράφεται εδώ https://el.player.fm/legal.
Good morning from OWITH.ai: the podcast that gives you only what's important to hear in AI and tech world.In March 2024, OpenAI CEO Sam Altman hinted at the upcoming GPT-5 model, but it fell short of expectations. The scaling laws that AI companies relied on for improvement are now breaking down, as seen with the underwhelming performance of the Orion model. Cognitive scientist Gary Marcus had previously warned about the limitations of deep learning, and his concerns are now being validated. OpenAI's new direction with the O1 model, focusing on real-time pondering and improved reasoning, represents a shift away from traditional GPT models. The success of this new paradigm remains uncertain, as it may require more time and resources to generate intelligent responses. In March 2025, the release of Orion leads to disappointment among users, shareholders, and critics, signaling the potential end of generative AI.Lina Khan's departure as chair of the Federal Trade Commission under the new Trump administration may not be universally celebrated in the tech industry. Trump's impact on tech regulation is uncertain, with potential changes in the crypto industry and state-level regulations playing a significant role. Despite speculation about increased M&A activity in the tech industry post-Khan, the actual impact of her antitrust lawsuits on VC portfolios is questioned. The battles over technology oversight are likely to occur at the state and municipal level over the next four years. Recent venture deals include companies like Parker and UnifyApps raising significant funding rounds. The future of tech regulation under Trump remains uncertain, with potential changes in the crypto industry and state-level regulations playing a significant role in shaping the landscape.The US has imposed export restrictions on advanced chips to China, leading Taiwan's TSMC to suspend chip shipments. Swiggy's IPO in India is oversubscribed, despite the company not yet being profitable. Chegg is facing a decline as students turn to ChatGPT for homework help. Visa and Tencent are launching a palm-based authentication system for payments in Singapore. Apple is expected to focus on wearables for future growth, moving away from the iPhone. Amazon is changing its physical stores, and an AI painting sold for a high price at auction.The email is a summary of the 2024 Fortune Global Forum in New York City, featuring speakers such as Mike Pompeo, Leon Panetta, and Boris Johnson. Amazon CEO Andy Jassy plans to reduce bureaucracy by increasing the ratio of employees to managers. Market research suggests bullish market performance could extend through the 2030s with Republican control of the government.The text discusses the increasing alignment between Silicon Valley and the Pentagon, leading to the militarization of AI and the development of dystopian applications such as automated weapons. The U.S. is leading in drone technology, with big tech companies partnering with defense agencies to compete in the global AI arms race.The text discusses the possibility that the core hypothesis driving generative AI, particularly in terms of scaling, may be incorrect. The author argues that deep learning may not improve with scale for tasks such as reasoning and handling edge cases in open-ended environments. Recent evidence suggests that pure scaling may have reached a point of diminishing returns.

Thanks for listening! Follow us on Twitter, Instagram and Linkedin

  continue reading

82 επεισόδια

Todos os episódios

×
 
Loading …

Καλώς ήλθατε στο Player FM!

Το FM Player σαρώνει τον ιστό για podcasts υψηλής ποιότητας για να απολαύσετε αυτή τη στιγμή. Είναι η καλύτερη εφαρμογή podcast και λειτουργεί σε Android, iPhone και στον ιστό. Εγγραφή για συγχρονισμό συνδρομών σε όλες τις συσκευές.

 

Οδηγός γρήγορης αναφοράς