💾 Sluggish silicon

🧵 In today's edition:

• 🌪️ GenAI storms the enterprise

• 💾 Sluggish silicon

• 🗣️ Voices of AI

• 🤑 AI Fundraising News

• 🗞️ AI Quick-Bytes

• 🐦 Tweet Post of the Day

(Sponsored) 100DaysOfNoCode & 100DaysOfAI

Learn No-Code and AI skills by committing to the 100DaysOfNoCode or 100DaysOfAI challenge.

Receive a 30-minute bite-sized lesson straight to your inbox every day. It’s free, fun and effective! Both challenges start on April 1st, 2024 ⚡️

Start learning No-Code or AI skills today!

🌪️ GenAI storms the enterprise

In 2023, GenAI companions and image makers took the internet by storm.

But here's the secret: going forward, the substantial profits will come from the enterprise sector.

A16z spoke to and surveyed Fortune 500 leaders who fully embraced GenAI and learned some interesting things.

Finding 1: Budgets are significantly increasing.

• On average, budgets to deploy GenAI features for 2024 have tripled.

• Last year's average spend of $7 million per company was just the beginning.

• One company increased its GenAI investment eightfold after achieving 90% cost savings using an LLM for customer service.

Finding 2: Open-source models are gaining popularity.

• Enterprises are aiming for a 50/50 split with closed source in 2024, with 46% of survey respondents stating they prefer or strongly prefer open-source models.

• Why? For control, customization potential, and to avoid vendor lock-ins as the model’s performance aligns.

• While cost is a consideration, these leaders are more concerned about protecting proprietary data and understanding why models produce certain results.

• Privacy and ethical issues persist, limiting fully customer-facing applications for now.

Finding 3: Productivity tools are trending.

• Enterprises are all in on internal productivity boosters like coding copilots, knowledge management, and summarization.

• One executive commented, "getting an accurate answer is worth the money,” and they are 100% right.

Overall: The billion-dollar consumption of GenAI in 2023 was just the prelude.

A16z predicts that the enterprise model API and fine-tuning market will exceed a $5 billion run rate by the end of the year.

So what does this mean? Developing tools to tap into this trend will create the next generation of AI billionaires 💰️ 

Read more: a16z

💾 Sluggish silicon

In a groundbreaking study, scientists have developed a microchip that uses photons (light) instead of electrons (electricity) for rapid vector-matrix multiplication - a key component in training neural networks like ChatGPT.

Remarkably, the creators assert that their light-based silicon solution can be produced using existing chip manufacturing lines without extensive retooling.

The potential is staggering - this technology could greatly improve the training and efficiency of AI systems, especially given the rapidly increasing demand for these powerful applications. 

One of the lead researchers also suggested that these photonic chips could be incorporated into next-generation GPUs as co-processors.

Just imagine - a system that can train advanced LLMs at a fraction of the energy cost. 

If this is really as revolutionary as it sounds, AGI might literally be one photonic breakthrough away ⚡

Read more: Live Science

🗣️ Voices of AI

Imagine Beethoven teaching you how to play the piano, or our favorite video-game character talking to you IRL.

Now stop imagining it because it has just become a reality with Character AI's Voice feature.

Here’s the TL;DR:

• The app now allows AI to embrace expressive vocal styles from a growing library of voices.

• Plus, it enables users to create and share custom voice packs by uploading audio clips or recording directly into the app (black mirror, anyone?).

The best part? This latest update gives every user free access to Voice - a huge step towards Character AI's broader vision of rich, multimodal interaction. Although it's currently English-only, the developers promise to add more languages for the global community.

The Voice rollout offers an interesting glimpse into the AI interface's future. As natural language abilities combine with emotive artificial voices, who knows what creative multimodal experiences the community will craft next?

Read more: Character AI blog

🤑 AI Fundraising News

• Stanhope raises £2.3M/$2.9M in funding for AI that teaches machines to “make human-like decisions”

• FundGuard raises $100M in a Series C funding round to focus on modernizing investment accounting operations through cloud-native technology and AI capabilities.

• Circular raises $10.5M in funding to provide access to a database of over 9,000 global suppliers with 50,000 PCR technical records using AI. The platform streamlines the searching, sourcing, vetting, and onboarding process for customers, providing faster results compared to traditional methods.

🗞️ AI Quick-Bytes

What else is going on?

• Nvidia debuts “more intelligent” AI-powered robots; truly self-reliant bots “further out.”

• Baidu shares rose after reports that Apple will use its AI services on China products.

• Verizon Business unveils AI-powered “Experiences Room” at Enterprise Connect.

• Radar-powered AI technology helps seniors age at home.

• What our shopping haul taught us about the promise of AI.

• Spotify is dabbling in online learning.

• Romanian state agency turns to AI to help farmers tap EU funds.

• Apple’s Cook tells China Forum AI is key for climate battle.

🐦 Tweet Post of the Day

Stability AI just launched a new model that massively improves code generation 🧑‍💻