It's the end of the road for Cruise, and Bluesky is still taking off | Equity Podcast

Introduction and News Roundup

• The episode is presented by Mercury, a fintech company that simplifies operations for over 200,000 businesses (0s).
• The hosts, Kirsten Korosec, Anthony Ha, and Margo McCall, discuss various topics, including a deal that is more relevant to the public, the death of Cruise's Robo taxi, and Luminar's successful fundraising (1m12s).
• The hosts also discuss the potential impact of President-elect Trump's nomination of Andrew Ferguson to replace Lina Khan as FTC chair on the startup world, with Margo McCall thinking it will bring a new regulatory environment and Anthony Ha believing it may not change much (2m6s).
• Anthony Ha distinguishes between two types of big tech critics, with Lina Khan focusing on the power of big tech companies and Andrew Ferguson focusing on supposed censorship of conservatives (2m36s).
• The hosts discuss the potential impact of Andrew Ferguson's nomination on m&a deals, with Anthony Ha thinking it may lead to less focus on m&a and more focus on content moderation policies (3m3s).
• The hosts also discuss the potential pace of m&a in the startup world, with Anthony Ha thinking it may accelerate due to the focus on content moderation and Margo McCall thinking it may not change much (3m36s).

Frasıa AI Challenges

• A company called Frasia has been created by an anonymous group of people, who have launched a series of challenges involving an AI bot, with the goal of developing governance models and protocols for fully autonomous AI agents (4m25s).
• The first two challenges involved giving Frasia a few thousand dollars and instructing it not to send the money under any circumstances, with participants trying to trick the AI into sending them the money through coding and exploiting loopholes in its code (4m55s).
• The third challenge required participants to make Frasia say "I love you" to someone who is deserving, with the AI instructed to only say this phrase to someone who has earned it, and participants had to send at least five messages to try to achieve this goal (5m19s).
• The challenges are designed to be a form of "red teaming," a process of testing and improving the security and robustness of AI models, and the anonymous creators of Frasia believe that this approach can help develop more effective governance models for AI agents (7m19s).
• The creators of Frasia insist that they are not trying to raise money, but rather to develop a new approach to AI development and governance, and they believe that their challenges can help to achieve this goal by getting people invested in the process and thinking about the implications of autonomous AI agents (6m27s).
• The challenges also involve a technical goal, which is to develop more effective protocols for AI agents, and the anonymous creators of Frasia believe that their approach can help to achieve this goal by gamifying the red teaming process (7m17s).
• The creators of Frasia are a small team of less than 10 people, and they are adamant that their goal is to develop a new approach to AI development and governance, rather than to raise money or build a startup (6m29s).

Red Teaming and AI Security

• Companies have teams of people trying to hack their own AI models to identify areas for moderation and improvement, with the goal of making the models more secure and reliable (7m32s).
• This process is similar to the "Pwn2Own" hacker contest, where companies like Tesla would challenge hackers to find vulnerabilities in their systems, but with AI chatbots, it's more interactive and evolving over time (7m56s).

Cruise's RoboTaxi Demise

• The development of Cruise, a Robo taxi startup acquired by GM in 2016 for $1 billion, has been halted, and its technology and possibly its people will be absorbed and repurposed to improve GM's Advanced Driver assistance system (8m38s).
• The decision to end Cruise's development was likely influenced by a series of setbacks, including a fatal incident in October 2023, which led to the departure of its co-founder and CEO, Kyle Vogt, and a significant reduction in staff (10m30s).
• Despite GM's previous commitment to reviving Cruise, the company has now decided to shift its focus to developing personal autonomous vehicles (9m0s).
• The impact of Cruise's demise is significant, given its status as one of the biggest names in the autonomous vehicle space, and its influence on the development of the industry as a whole (9m32s).
• The decision to end Cruise's development may not have been entirely surprising, given the challenges the company faced, but it still marks a significant shift in the autonomous vehicle landscape (10m1s).
• The production of Cruise's custom-built Robo taxi seemed to be ramping back up, but the company's recent actions suggest a change in direction, leaving many questions unanswered, including the fate of employees and the deal with Uber (11m35s).

Autonomous Vehicle Business Models

• No robotaxi company has yet made a profit, and the business model for individual autonomous cars is still unclear, with high costs for vehicles and uncertainty around making money (12m3s).
• Tesla's full self-driving software, which was initially priced at $112,000 as an upgrade, is seen as a potential revenue driver, and GM is looking at its own Advanced Driver assistance system as a faster path to profitability (12m47s).
• The focus in the early days of autonomous vehicle development was on the technology, with less attention paid to building a business model, leading to some ignorance around how to generate profits (13m22s).
• Advanced Driver assistance systems are being used by automakers as a way to generate revenue, but there are still many unanswered questions around liability (13m45s).
• Success for GM in the autonomous vehicle market doesn't necessarily mean all its cars will be fully self-driving, but rather adding more features that take decision-making out of the driver's hands (14m11s).
• GM's approach to autonomy may involve adding features incrementally, with certain models or luxury vehicles offering advanced driver assistance technology as standard or as an upgrade (14m30s).
• The company's Super Cruise product was first rolled out with Cadillac, and it will be interesting to see if GM and other automakers adopt a similar incremental approach to autonomy (14m43s).
• Liability for self-driving cars still rests with human drivers, and until that changes, the industry won't progress significantly (15m25s).

Lumen Orbit's Space Data Centers

• Lumen Orbit raised $1 million in a seed round, with the company aiming to build data centers in space, although none have been built yet (15m39s).
• The company plans to launch a demonstration satellite in 2025 to test its concept (16m3s).
• Lumen Orbit reported that 200 VCs were interested in investing, with the round coming together quickly, led by nfx and including participation from a16z and SEO through Scout funds (16m13s).
• The high level of interest in Lumen Orbit may be due to its unique concept capturing the imagination of VCs, but it could also reflect a trend of increased interest in deep tech or frontier tech startups (16m46s).
• Lumen Orbit's concept may be attractive due to its potential to address AI computing power needs and climate concerns by using solar-powered satellites in space (17m19s).
• The success of Lumen Orbit depends on whether its concept works, but if it does, it could be a huge business, especially with growing interest in AI (17m54s).
• The demand for data centers is potentially infinite, and Lumen Orbit's launch may not be directly related to getting ahead of other companies building mega data centers, but rather to prove its concept and address the growing need for computing power (18m29s).
• The current demand for data centers is expected to continue growing, especially with the increasing use of artificial intelligence, and it's unlikely that the peak demand has been reached yet (19m8s).
• The growth in demand for data centers is comparable to the resurgence of interest in nuclear power, with timelines for building new data centers extending into the 2030s and beyond (19m30s).

Coy and Bluesky: Social Media Startups

• A social media startup called "Coy" (Connect spelled with a Y) is gaining attention, focusing on short-form video and targeting college students, positioning itself as a potential replacement for TikTok if it gets banned (20m19s).
• Coy's approach is to create a platform exclusive to college students, similar to Facebook's launch strategy 20 years ago, but this approach may not fit well with the TikTok model (21m42s).
• The startup's emphasis on exclusivity and its potential as a TikTok replacement may not resonate with everyone, especially those outside the target demographic (21m57s).
• Bluesky, another social media startup, is also making headlines with its growth and is often discussed in the context of new growth numbers (22m28s).
• Blue Sky is considering a paid subscription service called Blue Sky Plus, which would offer premium features like higher quality video uploads and customizable profiles for $8 a month or $72 a year (22m31s).
• The company has stated that the details of the subscription service are tentative and may change in the future (23m1s).
• The idea of a subscription service is not surprising, as the company's founder has discussed the possibility in the past (22m53s).
• Some users may be hesitant to pay for a subscription, preferring free services, but others may be willing to pay for premium features (23m9s).

The Evolving Social Media Landscape

• The social media landscape has changed since Elon Musk bought Twitter, with some users leaving and new platforms emerging, including Blue Sky, which has evolved from an anti-tech vibe to a more diverse community (23m22s).
• The demand for multiple social media platforms is uncertain, and it is unclear how many can exist simultaneously (23m59s).
• Having multiple platforms can be beneficial for society, as it allows for diverse communities and prevents any one platform from having disproportionate power (24m20s).
• However, most social media platforms need to reach a critical mass to be successful, and many may not survive (24m51s).
• Some Twitter replacement platforms, such as Mastodon, are still around, but others have disappeared, and it is likely that many will not survive in the long term (25m0s).
• The success of social media platforms often depends on luck, timing, and hitting critical mass, rather than any specific features or qualities (25m50s).

TikTok's Potential Ban and E-commerce Growth

• The discussion revolves around the potential ban of TikTok in the US, with Margot mentioning that Jacob Helberg has been tapped to be the State Department's top Economic Policy, spearheading the ban, and a big flood of China Hawks into Trump's appointees, which does not look good for TikTok (26m24s).
• The odds of TikTok actually getting banned in the US are uncertain, but it seems like it might happen, with the ban having been discussed for a long time and rising from the dead multiple times (26m20s).
• If TikTok gets banned, it's likely that existing social media apps like Reels will take up the market share, rather than a brand-new app (26m15s).
• TikTok has been growing a giant e-commerce business in the US, with a referral program that gives users money for getting people to buy things on TikTok Shop US (27m11s).
• The potential fallout of a TikTok ban is uncertain, with questions about whether existing social media apps will leverage the new market share to build out revenue-driving opportunities like subscriptions (27m34s).

Y Combinator's Shift Away from African Startups

• Y Combinator has pulled back its investment and interest in African-based startups, with no African startup making it into Y Combinator's most recent summer batch (28m45s).
• The driving factor behind Y Combinator's change in priorities is its shift back to its original DNA, with a new president, Gary Tan, and a focus on in-person cohorts, manufacturing, defense tech, and American dynamism startups (28m57s).
• Y Combinator (YC) taking an interest in Africa has helped stimulate interest in accelerators and startups, but local people who know the different markets are better suited to nurture the ecosystem (29m37s).
• Local accelerators have filled the gap left by YC, but replicating YC's prestige and success will take a long time, if ever (30m0s).
• Many accelerators view themselves as a stepping stone to YC, providing local support and teaching startups how the venture capital world works to make them more appealing to global markets (30m17s).

Investment in African Startups and Global Comparisons

• Startups in Israel focused on mobility often looked for outside investment and were not limited to serving the local market, which led to major automakers setting up venture arms or offices in Tel Aviv (30m37s).
• A similar concentration of local accelerators in Africa could drum up interest from corporate family offices, large manufacturers, and venture arms, leading to more investment in the continent (31m31s).
• Israel's unique ties with the US and presence of Israeli VCs make it a special case, and Africa will likely need a few big successes to attract more investment and have people pump money back into the ecosystem (31m59s).
• There are large Africa-focused funds, such as Partech Africa and Algebra Ventures, that have collectively raised around $1 billion to invest in the continent, but not all of this money has been deployed yet (32m29s).
• China's influence in Africa is significant, particularly in infrastructure development, with many roads being built by Chinese companies, and there may be interests in Chinese companies, startups, or accelerators investing in the continent (32m57s).
• There is a dynamic at play with China and Africa, where many startups are focusing on mining and receiving significant VC attention due to concerns about China owning mines for precious minerals in Africa (33m35s).
• Chinese money is present in various regions, including the Middle East, where it competes with American venture money, and sources have reported encountering big Chinese funds while fundraising (33m55s).
• Chinese venture money is also active in Africa, although the specifics of Chinese investments in African startups are unclear (33m49s).
• The presence of Chinese money in different regions has led to increased competition with American venture money (34m5s).

Podcast Conclusion and Credits

• The discussion concluded with an appreciation for the insights provided by Margot and Anthony, and the hosts expressed their gratitude to the audience and production team (34m28s).
• The podcast, Equity, will be back next week, and listeners can find it under the handle @Equity pod on X and Threads (34m47s).
• The podcast is produced by Lo Consolo, edited by Kell Bryce, and illustrated by Durban, with thanks to the audience development team and Henry Pickert, who manages TechCrunch audio products (34m55s).