Tech Companies' AI Investment Strategies

🚨 A tech CEO-turned-investor, Jill (Greenberg) Chase oversees AI and data infrastructure investments for CapitalG, Alphabet Inc.’s independent growth fund. She joins us for VC Wednesdays. 🚨 ✒️ How would you describe your thesis? There are four categories in AI that you could look at. One is the models themselves. There are full-stack companies, which are building their own models to power end applications. The third is wrappers on top of foundational models. And the fourth is infrastructure to either build or to use LLMs. My focus area is full-stack applications, which can have interesting tech moats because you're building your own model and it’s often more cost-effective than using a massive model to power an end use case. My second area of focus is the infrastructure layer. ✒️ What’s the No.1 thing that you look for before making an investment? The number-one thing I look for is an outlier metric – something that is so unique that you have to pay attention to it. That can be an outlier team, cohort data, product love or an outlier market that's so massive that it's begging for disruption. Beyond that, the questions to ask are: Is the team exceptional? Is the product loved? And is the end market big enough to support a big outcome? ✒️ What’s a recent example? I led the series A round in Magic. They do not have a product yet, but they’re building an AI programmer that can enable actual code generation, versus code auto-complete, which is in the market today. Their outlier metric is a novel model architecture that’s different from transformers and allows for long context windows, which will enable a truly game-changing product in AI-generated code. ✒️ What’s your top AI prediction for 2024? We saw incredible progress on the actual technology for LLMs last year, but we were still limited by the cost to deploy them at scale and their performance in terms of hallucinations, context windows and multi-step reasoning. So you got fits and starts of interesting use cases, but very few examples of enduring value both for end customers and for sustainable business models. My prediction for 2024 is that we're actually going to start to unpack some of the multi-step reasoning and the context windows, which will enable real value and use cases. ✒️ How can AI startups build sustainable business models? Companies need to account for inference. It's extremely expensive pinging an API for some of these LLMs. So if you have a high-volume use case, it's going to be very hard to make the unit economics work. It's unlikely that you're going to be able to pass that cost on to your end customer, so you have to be really thoughtful about either finding ways to leverage smaller models that are fine-tuned and bring that inference cost down, or figuring out ways to provide more value via integrating LLMs into your product and passing along that price to your end customer. #VCWednesdays #vc #venturecapital #startups #TechonLinkedIn

How 300 Software Startup Execs ($10M-$1B+ Revenue) Are Actually Using AI in 2025 Just read through a 67-page report surveying executives at high-growth software companies like Cursor, ElevenLabs, and Sierra. Here's what caught my attention: The Model Race: OpenAI Still Leads, But Claude is Closing In -OpenAI remains the top enterprise model provider, but Claude has secured a solid second place. The enterprise AI landscape is becoming a two-horse race at the top. Where the Money Really Goes -Here's the surprise: companies are spending more on data storage, processing, and AI infrastructure than on inference and training. But AI talent? That's still by far the biggest expense line item. The Development Stack -The report reveals which tools are actually being used to build AI products (not just market hype). The real development toolkit looks different from what you might expect. Scale = Serious Investment -Companies hitting around $500M in revenue are dropping roughly $100M annually across training, inference, data storage, and processing. That's 20% of revenue going to AI infrastructure. 2025: The Agent Economy Arrives -Nearly 90% of high-growth startups are either actively deploying or experimenting with AI agents. We're not talking about chatbots anymore—these are autonomous systems handling real business processes. What This Means -The AI implementation gap between early adopters and everyone else is widening fast. Companies that figured out their AI strategy in 2024 are now scaling infrastructure investments that smaller players can't match. -The enterprise AI market is consolidating around proven providers while operational costs are shifting toward data and infrastructure. If you're building in this space, your competition isn't just other startups—it's companies with $100M AI budgets. What's your take? Are we seeing the formation of an AI infrastructure moat that's going to be hard for newcomers to cross?

AI Is Making “Impossible” Deep Tech Ventures Merely “Extremely Hard” — Here’s How 🚀 Deep tech, biotech, and pharma startups tackle humanity’s greatest challenges: curing diseases, fighting superbugs, creating sustainable materials, and powering new industries.  But the path is brutal—decade-long timelines, hundreds of millions in funding, and massive scientific risk. For every success, countless ventures fail in the “valley of death.”  AI is changing the game by compressing time and slashing capital needs across the board:   • AI-driven target identification finds novel drugs faster than years of manual research (think Insilico Medicine).   • Generative AI designs drug candidates rapidly, cutting trial-and-error chemistry drastically.   • AI optimizes clinical trials by predicting outcomes and speeding patient recruitment.   • Digital twins simulate advanced materials and devices, replacing costly physical prototypes.   • Automated R&D platforms run experiments 24/7, accelerating discovery cycles.   • AI-enabled manufacturing boosts yield and scale-up reliability for biologics and semiconductors.  According to our latest analysis, even a 10% reduction in perceived scientific risk from AI “de-risks” investments, unlocking more capital at better valuations. Plus, if AI halves development time, venture funds can double their shots on goal—dramatically increasing the number of breakthroughs.  This is the AI-Driven Enterprise (AIDE) model: AI at a company’s core, transforming how ventures are built from day one.  What if your startup or fund adopted this mindset now? How much faster and smarter could humanity solve its biggest problems?  👇 Share your thoughts or examples of AI reshaping deep tech ventures in the comments!  #DeepTech #Biotech #AIInnovation

The AI race is presently being fuelled by venture capital funding. According to PitchBook, over $29.1 billion was invested through different venture capital arrangements across over 700 generative AI deals in 2023. And it appears that 2024 will clearly surpass those numbers. Having already invested $1.25bn in Anthropic, Amazon is making its largest external investment by investing another $2.75 billion in the AI company. As part of the agreement, Anthropic said it will use AWS as its primary cloud provider. It will also use Amazon chips to train, build, and deploy its foundation models. Amazon’s move is the latest in a spending blitz among cloud providers to stay ahead in the AI race. And that is even as generative AI has gone mainstream. OpenAI has said more than 92% of Fortune 500 companies have adopted the platform, spanning industries such as financial services, legal applications, and education. Anthropic, valued at $18.4 billion, recently released Claude 3, its newest suite of AI models that it says are its fastest and most powerful yet. The company said the most capable of its new models outperformed OpenAI’s GPT-4 and Google’s Gemini Ultra on industry benchmark tests, such as undergraduate level knowledge, graduate level reasoning, and basic mathematics. Google has also backed Anthropic with its own deal for Google Cloud. It agreed to invest up to $2 billion in Anthropic, comprising a $500 million cash infusion, with another $1.5 billion to be invested over time. That's $6bn into Anthropic from 2 Big Tech companies. This corporate VC approach by Amazon and Google is in step with Microsoft's investments in OpenAI. $13bn has been invested so far by Microsoft and more could still be in the works. Meta has already indicated it is spending "billions" investing in Nvidia and is projected to spend $94-$99 billion this year on AI projects and investments. The influx of corporate venture capital, exemplified by these significant investments, is reshaping the landscape of the AI industry. With billions poured into companies like Anthropic, equipped with cutting-edge AI models outperforming industry benchmarks, the race for AI dominance intensifies. This strategic move not only fuels innovation but also solidifies partnerships, such as Anthropic's adoption of AWS as its primary cloud provider, further propelling the evolution and adoption of AI technologies across various sectors. As Big Tech giants join the fray with substantial investments in AI, it underscores the pivotal role corporate VC and venture capital as a whole, plays in shaping the future trajectory of artificial intelligence. And this may just be the beginning.