Competition in AI is heating up. You're poised to benefit.

Hi, Friends —

✰ Welcome to [our digital disco]! This week’s newsletter will operate a bit differently than usual. I will dive into recent developments in AI and more specifically, the future of the tools we already interact with daily. The competition is accelerating product development, meaning that our day-to-day work will shift drastically in the coming months.

✰ Keep an eye out for an upcoming exploration of the future of our internet use, plus advice for leveraging these tools to become more potent now. You can also check out last week’s newsletter here.

The firms giving OpenAI a run for its money

A number of competitors — most significantly, Google and Anthropic — are challenging OpenAI’s primacy in the AI-powered chatbot and productivity market. Others such as Amazon and Meta are appearing at the scene, revamping their teams and investments to double-down on AI-powered features.

For many, OpenAI's GPT models became the gold standard for large language models (LLMs), particularly with the explosive popularity of ChatGPT and GPT-4. The excitement about these AI tools kickstarted an arms race in AI. Big Tech and solopreneurs alike are sprinting to the scene, fueled by the investors pouring billions into their ideas.

• Why does it matter? AI has exploded in popularity over the last few months, largely underpinned by technological advancements and widespread accessibility. While the market has been hot since the arrival of ChatGPT, it’s worth noting that a number of firms are proving fierce competition. Much has happened over the past three weeks, enough to commit an entire newsletter to the competitive landscape.

• Pros: This fierce race to innovate is poised to bring significant benefits to users like you. The AI competition fuels a continuous cycle of improvement, as companies relentlessly work to refine their AI models, address biases, and deliver regular updates. With each company striving to outdo the other, you can expect access to advanced functionalities, which may revolutionize your workflows and increase efficiency.

• Cons: It’s vital to acknowledge the potential drawbacks and daunting challenges that lie ahead. Perhaps the biggest problem is that of hallucination — LLMs have a tendency to fabricate information. As companies scramble for dominance, they may compromise quality control measures, leading to the spread of harmful content and misinformation. Ethical considerations, such as user privacy and data security, may also take a backseat as companies prioritize their competitive edge.

Let me give you a quick rundown on key developments in the competition, and the tools you can leverage to power your productivity.

• Anthropic’s Claude now has the largest context window of currently-available models — meaning it can intake vastly more information to generate highly-specific responses. Claude can process enough information as an entire novel in under a minute. This gives it has an incredible ability to digest and synthesize materials such as legislation and dense business documentation. Anthropic also boasts less harmful and more accurate outputs compared to those of its competitors, due to a commitment to safety, equality, and human rights, among other principles.

• Microsoft publicly launched its AI-powered search engine Bing earlier this month, and also revealed massive upgrades — including an Actions feature that allows users to complete tasks without navigating among multiple sites. (Think: Finding and booking a restaurant res in one fell swoop). Since making a $10B investment in OpenAI in January, Microsoft tools have been receiving ongoing facelifts powered by GPT-4 — prompting a fire drill within Google to maintain its online search primacy.

• Google has emerged as a fierce competitor in the AI scene, and is lined up to run directly against incumbents in the generative AI space. As announced on Wednesday at its annual I/O keynote conference, Google is integrating AI into virtually all of its products. Some highlights:

  • “AI snapshots,” or AI-powered search results, generate mini reports designed to answer one question and anticipate the next.

  • Google Bard, the chatbot that rivals ChatGPT and Bing, is now available to users in 180+ countries. Unlike ChatGPT, Bard has real-time access to the internet and extensive translation capabilities.

  • Duet AI, Google’s answer to Microsoft Copilot, embeds generative AI into Docs, Slides, Gmail and more (imagine auto-generated slides, emails, etc).

Finally, some other key firms are rumored to be joining the bandwagon:

• Amazon is looking to reimagine search through new conversational experiences — i.e., embedding generative AI into its search features similar to Bing and Google — according to recent job listings. The shift comes alongside news that Amazon is building a new LLM to power Alexa, a voice assistant which Amazon has historically struggled to make profitable.

• Apple is also considering LLM-based improvements to Siri. The voice assistant remains far behind products such as ChatGPT and is often seen as a slow, unreliable tool. Updating Siri to operate at a competitive speed and quality would require Apple’s significant investment in LLMs, or perhaps a strategic partnership with a leading firm (similar to the actions taken by Microsoft).

• Samsung has teamed up with tech giant Naver (Korea’s top search engine provider) to develop an in-house AI tool that rivals ChatGPT. The firms seek to improve productivity while addressing their security concerns of using external APIs. The move is particularly interesting because this new tool feeds into the partners’ larger AI ecosystem. Samsung and Naver will soon unveil next-generation AI chips to power the technology and compete against those of incumbent Nvidia’s graphics processing unit (GPU).

Of course, there are many, many more developing tools and models. Hugging Face recently joined the ChatGPT competitors with open-source alternative HuggingChat. Baidu is leading the generative AI and chatbot scene in China. Meta has poured billions into AI development (and has a model that may provide the basis for a ChatGPT alternative). Cohere, Stability AI, and others also have models that are gaining popularity.

✰ Next in [our digital disco], I’ll dive into how spaces like search engines and productivity tools are being transformed by AI — and my recommendations for leveraging them in your own work.

Snacktime

📓 Reading: The world needs an international agency for artificial intelligence, say AI experts Gary Marcus and Anka Reuel. I’ve long argued for an organization with this goal, and it’s been fascinating to see this discussion emerge to the public square.

♬ Listening to: A great episode on OpenAI, its activities in trust and safety, and more. The Trust in Tech podcast is run by the Integrity Institute, a think tank focused on protecting the social internet.

✿ As always — any and all feedback is welcome! In the meantime: give someone a hug and say an ‘I love you’ this week. Make the world a little happier.