Justin's AI Newsletter

The most interesting content, news, and insights in AI each week.

--

Hey friends, happy June! This is officially "AI Summer". The pace of AI innovation has continued to speed up, and I think we'll continue to see exciting developments for the rest of the year.

Just imagine where things will be on June 1st, 2024...

Let's dive in.

---

🛣️ The OpenAI Roadmap

Sam Altman sat down and explained some of what is happening at OpenAI. It gives some clarity into what we can expect to see over the next couple months.

• They’re heavily limited by GPU’s. This explains why things like text-to-image and 32k context window hasn’t been rolled out to everyone.
• They’re working on larger context windows (possibly up to 1M tokens!) If you want to know what that means, read my post here.
• GPT-4 will become cheaper & faster.
• They’re still figuring out plugins. Eventually, plugins will be released via API, and anyone will have access to plugins.

One of the best things about OpenAI is their transparency. They are not afraid to let everyone know what’s going on, which is a powerful trait for an AI company.

---

📊 ChatGPT Stats (You're still early...)

This past week, we saw some new research about how often people are using ChatGPT. Honestly, some pretty interesting stats…

• Only 14% of U.S. adults have tried ChatGPT.
• ~40% of adults who have tried ChatGPT say it has been “somewhat useful”
• Only 2% of ChatGPT users find it “Extremely Useful”

In other words, we’re still in the early days. As someone who uses ChatGPT daily, I know with full confidence that it is “extremely useful”. The fact that only 2% of people share this opinion reveals the challenges surrounding user education and implementation.

The need to “perfectly prompt” ChatGPT to obtain ideal output is the problem. As time goes on, we can anticipate the development of new products that will make it easier for the average person to experience the full potential of ChatGPT.

---

😂 An incredible SEO Hack...

The other day, I saw a Thai food restaurant named "Thai Food Near Me".

It's an incredible SEO hack - if you search for "Thai Food Near Me" on Google, it's the top result.

I'm guessing it's also a great LLMO (large language model optimized) hack as well. 😂

---

💡 This Week's Recommendations

​US judge orders lawyers to sign AI pledge, warning 'they make stuff up': A Texas federal judge has ordered lawyers to confirm that they did not use AI to draft their filings without human review, citing the possibility of inaccuracies and biases. This follows another federal judge in Manhattan threatening sanctions against a lawyer for using an AI-generated citation to a bogus case. Hallucinations are in the wild…

​frontiers, startups, 2023++: Shyamal is a GTM leader at OpenAI. He breaks down his thoughts about AI startups as it relates to the future of AI. He shares some interesting frameworks, a good read if you’re building in this space…

​They Plugged GPT-4 Into Minecraft—and Unearthed New Potential for AI: Nvidia released a GPT-4 AI-agent into the simulated world of Minecraft. It showed how this Agent was able to create it's own objectives and actually perform well in this world. The article suggests that using language models in this way could automate many routine office tasks, potentially one of the technology’s biggest economic impacts.

​AI experts challenge ‘doomer’ narrative, including ‘extinction risk’ claims: AI experts suggest that the focus on the existential risk of runaway AGI detracts from current risks such as bias, misinformation, high-risk applications, and cybersecurity. While some worry about low-probability events in the long term, most are not highly concerned about existential risk. These AI experts believe that organizations warning of existential risk obtain funding by convincing donors that AI existential risk is a real and present danger, but focusing solely on these risks will negatively affect addressing current harms.

---

Thanks for tuning in,

Justin

​Justin's Full Content Library | Previous Newsletters​

113 Cherry St #92768, Seattle, WA 98104-2205
​Unsubscribe · Preferences​