OpenAI's Historic Funding Round, AI Dominates VC Deals | Bloomberg Technology

04 Oct 2024 (10 days ago)
OpenAI's Historic Funding Round, AI Dominates VC Deals | Bloomberg Technology

OpenAI's Funding and Valuation

  • OpenAI has completed a deal to raise $6 billion in new funding, giving the AI company a $157 billion valuation and bolstering its efforts to build the world's leading generative AI technology (1m2s).
  • The company also announced a $4 billion credit line from banks around the world, giving access to more than $10 billion war chest (1m22s).
  • OpenAI needs the funding because its cutting-edge model requires significant investment to trade and build, with costs running into hundreds of millions of dollars (1m37s).
  • The company is in an expensive business, with costs including chips, staff, and other expenses, despite bringing in revenue (1m44s).
  • The future growth model for OpenAI is based on businesses using their products, with the company hoping that businesses will fine-tune the models, use them in different parts of their workflow, and find them useful enough to keep paying for them (2m41s).
  • OpenAI has seen uptake from businesses, with over one million business users of ChatGPT, and 250 million weekly users (3m59s).
  • The company's valuation is based on its potential for future growth, with investors willing to wait for a long-term return (4m42s).
  • The funding round and credit facility were backed by substantial banks, including Goldman Sachs, J.P. Morgan, Santander, and Wells Fargo (2m29s).
  • OpenAI's products span different areas, including an educational product, smaller business product, and more enterprise products (4m12s).
  • The company's future impact and influence will be discussed in a future interview (5m4s).

Investigation into SAP's Pricing Practices

  • An investigation is looking into companies, including SAP, for a potential scheme to fix and inflate prices in sales of SAP technology to the U.S. government, with the scope of the investigation expanding to include sales to as many as 94 civilian government agencies, in addition to the Department of Defense (5m33s).
  • The companies involved, including SAP and Carahsoft, have made statements, with SAP saying there was no requirement to report, and Carahsoft also providing a statement, as they play a historic role in selling software to the U.S. government, with Carahsoft handling the complicated process of navigating regulatory hurdles (7m16s).

AI's Impact on Jobs and the Economy

  • The SoftBank founder has envisioned a near future where AI technology would run entire households, with artificial general intelligence (AGI) arriving in the next two to three years, but not everyone is bullish on AI, particularly when it comes to jobs (8m43s).
  • A renowned MIT professor says that by his calculation, only a small percent of all jobs, around 5%, is ripe to be taken over or heavily aided by AI over the next decade, with this thesis based on the current capabilities of AI, which is far from performing tasks that involve a heavy level of interaction with the physical world (9m49s).
  • The professor's skepticism is based on the limitations of AI in tasks such as construction work, manufacturing work, and carpentry, which may become more feasible if AI becomes more advanced, reliable, and integrated with better robots (10m12s).
  • The impact of AI on the economy and jobs will not be significant in the next five to 10 years, with only around 5% of jobs being affected, as AI is not yet capable of performing complex social tasks (10m41s).
  • The idea that AI technology is exponential and will lead to significant breakthroughs in the near future is uncertain, and the most optimistic views in the industry may not hold water (11m17s).
  • There is huge uncertainty in forecasting AI's impact, and one source of uncertainty is what will happen with AI in the next several years (11m47s).
  • Doubling the amount of data and computer power will not necessarily lead to double the capabilities of AI, as higher-quality data is needed for AI to perform tasks like those of a carpenter (12m8s).
  • The current dominant architecture of AI has limitations in terms of reliability and high-level reasoning, which is why significant improvements in AI capabilities are unlikely in the next five to 10 years (12m39s).
  • The belief that AI will only improve gradually is based on historical norms of other waves of digital technology that promised automation and productivity improvements but delivered gradual changes (13m23s).
  • Advances in AI have been seen over the last two and a half years, but they have not exceeded the basic capabilities of GPT 4, and future advances are unlikely to quickly change the way tasks are done in the economy (13m43s).
  • It is unlikely that certain occupations will completely disappear in the next five years, but rather, AI will augment human capabilities and boost productivity (14m12s).
  • The future of AI is likely to be a tool for workers to boost their productivity, with some areas already seeing phenomenal improvements, such as writing simple subroutines in common programming languages (14m51s).
  • AI can provide help to decision-makers in certain areas, but its impact on the whole economy will depend on its ability to help a wide range of workers, including those in construction, transportation, and teaching (15m16s).
  • AI models are not yet reliable or reasonable enough to achieve tasks such as journalism, but they may be able to do so in the future (15m31s).
  • AI is leading to a boost in jobs for people with AI fluency, such as AI programmers, integrators, and other AI-related service activities (15m47s).
  • Research from a previous wave of AI in the late 2010s did not show any other job category getting a boost in terms of hiring, and some jobs, such as I.T. security, may decline due to AI tools (16m8s).

The OURA 4 Ring and the Future of Wearable Technology

  • The OURA 4 ring has debuted, featuring a sleeker design and improved smart sensing technology that dynamically adjusts to the user's body (18m35s).
  • The OURA 4 ring is seen as a complementary device to smartwatches, offering an eight-day battery life and seamless tracking of the user's physiology without notifications (19m21s).
  • The OURA 4 ring is compatible with other devices, and the company is working with other businesses, such as Meta, on integration and partnerships (20m25s).
  • A world is envisioned where a cloud of devices on the body, used in different contexts, can work together with a person's physiological state to provide input and guide other wearables, offering opportunities to explore with permanently worn health-oriented devices (20m48s).
  • The company's focus has been on technology and innovation, with a trajectory showing growth, and the company sees a lot of headroom to grow in the ring category, with people learning about the benefits of long-term health tracking (21m30s).
  • The ring category offers various benefits, including understanding the body, tracking sleep, and providing insights, even with an AI coach that helps make behavior changes, and there is a lot of headroom for growth in this category (21m50s).
  • Beyond the ring category, there are opportunities for interaction with other devices, and prototypes have been seen for a locket and designer rings, with a product introduced with Gucci that looks like jewelry and fits into one's life (22m10s).
  • The future is bright for rings, with lots of headroom to grow the business, and people being aware of their health and understanding how to manage it is a big opportunity, with the goal that everyone on the planet should have one of these devices (22m25s).

Market Update: Chinese Tech ETF, NVIDIA, and Chipmakers

  • A new ETF launches today, focusing on Chinese tech stocks, and the NASDAQ 100 is flat, with modest outperformance in semiconductors, and Bitcoin has come under pressure (22m48s).
  • NVIDIA is still the name people are talking about, worth more than $3 trillion, with insiders selling more than $1.8 billion in shares so far this year, but the CEO's sales were preplanned and correlated with the stock's drawdown (24m9s).
  • The CEO's sales do not necessarily indicate growth concerns, as the stock is still up close to 150% year to date, and there are supposed to be additional sales in the future, which may impact the stock's performance (24m54s).
  • Evercore made a call for chipmakers, predicting a rebound in the fourth quarter of this year, citing a classic story of wealth creation from stock compensation in Silicon Valley (25m29s).
  • The sales of shares by executive directors have raised questions about whether they inspire confidence in the stock, with some portfolio managers arguing it's not just due to growth concerns (25m58s).
  • Round Hill has launched a new ETF that tracks an equal-weighted basket of Chinese tech companies, with a more targeted approach than other ETFs (27m2s).
  • The ETF, which has a 59 basis points fee, is trading today and showcases the demand of clients despite negative rhetoric around China (28m40s).

Venture Capital Market Update and AI Investments

  • Venture money is going towards AI startups, with Poolsides, a startup building AI for programmers, raising funds in a round led by Bain Capital (30m21s).
  • Venture deal-making has been slow, with investors becoming more selective and focusing on a handful of companies, with OpenAI's latest fundraising round being a notable example (30m56s).
  • The data as of the third quarter shows that the US venture market has invested around $37.5 billion, which is lower than the previous quarter, indicating a slow market (31m32s).
  • There is a lack of exciting investments outside of AI, with the majority of investments happening in San Francisco and the US, particularly in companies that require a significant amount of capital (31m55s).
  • The OpenAI deal shows the excitement around AI and technology, with investors like NVIDIA, Microsoft, and SoftBank looking at it as a potential source of huge returns (32m9s).
  • The inclusion of the OpenAI round in the data would make 2024 look much better, but it's essential to note that it's a single deal and not representative of the entire market (32m50s).
  • AI companies have higher valuations, larger deal sizes, and are more exciting and interesting, but this is not representative of the rest of the market (33m11s).
  • Exits are still a crucial factor in venture activity, and the lack of exits in the past two years has contributed to the slow market, with even IPOs in Q1 and Q2 unable to sustain the venture market (33m50s).
  • The slow market is expected to continue in the coming quarters, with limited partners having no way to use their cash for more investments or to recycle it back into venture (34m19s).
  • Lux Capital Partner, Dena, disagrees that there is nothing exciting outside of AI, citing intersections with sciences, physical intelligence, health tech, and Medicaid care delivery as areas of exciting investments (34m42s).
  • Dena notes that while AI is an exciting area, it's not the only area of exciting investments, and Lux Capital has been investing in AI for some time (35m6s).
  • Venture capital firms are still excited about investing in areas like AI, despite some companies struggling to raise early funding, with seasoned operators getting large mega rounds and companies across different stages and sectors continuing to raise capital (35m31s).
  • The big question is what the exit market looks like, with many looking at 2025 as the year for more exits, and venture capital firms having long lifecycles and patient investors who know companies take time (36m2s).
  • Venture capital firms like to invest in the impossible and turn it into the inevitable, looking for generational large companies that turn into very big outcomes (36m25s).
  • Some limited partners (LPs) may not be allowed to invest in competitors, which is a normal but not universal approach, and venture capital firms try to be careful not to invest in competitive companies and instead put resources behind their investments (36m56s).
  • The type of investor is different, with large mutual funds and companies like NVIDIA being players, and venture capital firms need to adapt to these new types of players (37m40s).
  • Having a large fund, over $1 billion, allows venture capital firms to invest in everything, construct a portfolio across different stages, and buy low and sell high (37m52s).
  • Venture capital firms can also start companies themselves, like Evolutionary Scale, which applies AI to life sciences and the fertility space (38m19s).
  • Companies that are not AI natives need to cope with the current environment by cutting costs, dialing in on revenue growth, and using AI to enable efficiency and better profit margins (38m45s).
  • All companies need to have an answer for how they will use AI to survive in the world, and some of the most exciting applications are in human-oriented companies that leverage community health workers and AI to deliver healthcare (39m16s).

Western Semiconductor Exports to Russia

  • The U.S. joined with European allies to impose export controls on Russia to deny the Russian military access to Western semiconductors and other technology after the invasion, and tightened those restrictions again starting last year (41m21s).
  • Despite these restrictions, there have been various reports of trade data about components, including semiconductors, getting through to Russia (41m42s).
  • An investigation found that Western components, specifically from Analog Devices, have been frequently found on the battlefield in Ukraine, highlighting Russia's dependence on Western technology to guide its missiles (42m2s).
  • Analog Devices regards this as an illicit diversion of their products and does not condone any of this trade (42m39s).
  • The complex supply chain for semiconductors, involving production in Southeast Asia and middlemen acting through Hong Kong using shell companies, has made it difficult to crack down on the trade (42m48s).
  • The U.S. has tried to enforce these controls through the Commerce Department, but it has proven difficult due to the global nature of the supply chain and the many layers semiconductors get traded through before reaching Russia (43m5s).
  • The story of Russian missiles using American chips is a must-read, titled "Russian Missiles, American Chips" (43m39s).

Overwhelmed by Endless Content?