ChatGPT Agent Doesn’t Cut It
Hello everyone. Happy Tuesday.
Today’s update may make some people deep in AI mania a little bit uncomfortable as it has to do with signs of trouble on the horizon. The latest example is found with OpenAI and AI agents. Let’s discuss.
ChatGPT Agent Doesn’t Cut It
Last week, OpenAI unveiled a general purpose AI agent.
Here’s TechCrunch:
“OpenAI is launching a new general purpose AI agent in ChatGPT, which the company says can complete a wide variety of computer-based tasks on behalf of users. OpenAI says the agent can automatically navigate a user’s calendar, generate editable presentations and slideshows, and run code.
The tool, called ChatGPT agent, combines several capabilities from OpenAI’s previous agentic tools, including Operator’s ability to click around on websites, as well as Deep Research’s ability to synthesize information from dozens of websites into a concise research report. OpenAI says users will be able to interact with the agent simply by prompting ChatGPT in natural language.
ChatGPT agent is rolling out on Thursday to subscribers to OpenAI’s Pro, Plus, and Team plans…
The launch of ChatGPT agent represents OpenAI’s boldest attempt yet to turn ChatGPT into an agentic product that can take actions and offload tasks for users, rather than just answering questions. In recent years, Silicon Valley companies including OpenAI, Google, and Perplexity have unveiled dozens of AI agents that have promised to do just that. However, these early version AI agents have proven to struggle with complex tasks, and they seem less compelling as products than the ultimate vision tech executives pitch around AI agents."
OpenAI told The Verge that agent is a result of the Operator and Deep Research teams being combined to produce a team of approximately two to three dozen people.
The basic idea on display here isn’t exactly new. Every prior attempt to figure out agents has led to dead ends, forgotten promises, and an overall yawn.
Not to bury the lede: This agent unveiling and launch was not good enough.
Become a member to continue reading today’s update. Already a member? Read the full update here.
An audio version of this update is available to members who have the podcast add-on attached to their membership.
Member Privileges and Benefits
Become an Above Avalon member and receive the following privileges and benefits:
Exclusive Analysis. Receive the Above Avalon Daily newsletter, widely-recognized as the leading daily newsletter dedicated to Apple. Now in its tenth year.
Archive Access. Access previous newsletters sent to members.
Member Forum Access. Access all channels in the Above Avalon forum in Discord.
Email Access. Receive timely responses from Neil to email inquiries.
Access to Add-ons. Customize a membership with the AVALON, Podcasts, Inside Orchard, and Financial Models add-ons.
Above Avalon Support. Play an active role in supporting Above Avalon as an independent voice and resource.
OpenAI’s Spring Update Event, GPT-4o, Apple Hardware at the OpenAI Event
Hello everyone.
Let's jump right into today's update.
OpenAI’s Spring Update Event
In what may become an annual tradition, OpenAI held a spring event to unveil ChatGPT and GPT-4 updates. GPT-4 is OpenAI’s multimodal large language model. The event took place just one day ahead of Google’s annual developer conference and about a month ahead of Apple's WWDC.
The livestreamed presentation (available here via YouTube), which appears to have taken place at OpenAI’s headquarters in San Francisco, had an audience consisting of two to three dozen people. Judging from the applause, the crowd was comprised of OpenAI employees. OpenAI CTO Mira Murati kicked things off and then guided the presentation along.
Murati anchored the event around three updates:
An Above Avalon membership is required to continue reading this update. Members can read the full update here. (Members: Daily Updates are accessible via the archive. If you haven’t logged into the archive before, fill out this form to receive an invite.)
Choose either a monthly or annual membership. Payment is hosted by MoonClerk and secured by Stripe. Apple Pay and other mobile payment options are accepted. After signup, use this link to update your payment information and membership status at any time. Contact me with any questions.
Contact me directly if you would like to purchase multiple subscriptions (five or more) for your team or company.
An audio version of the newsletter is available to members who have the podcast add-on attached to their membership. More information about the podcast add-on is found here. Special Inside Orchard bundle pricing is available for Above Avalon members. Additional membership customization is available via the Financial Models add-ons.
Google Releases Bard, Apple and Chatbots, DPReview to Shut Down (Daily Update)
Hello everyone.
We will kick off today's update with some chatbot discussion. Neil’s stance continues to be to approach chatbots with caution. We will talk more about that today. The discussion concludes with some additional comments about Amazon shutting down DPReview. There is more to that story than what we briefly alluded to yesterday.
Let's jump right in.
Google Releases Bard
“Google just announced that the company is releasing its ChatGPT competitor Bard. But chances are you won’t be able to access the product right away as the company is starting with a limited public rollout.
Users in the U.K. and the U.S. can head over to bard. google .com and join a waitlist. The company calls Bard an ‘early experiment that lets you collaborate with generative AI.’
Like OpenAI’s ChatGPT and Microsoft’s Bing chatbot, Bard is a chatbot based on a large language model. You can interact with Bard to ask questions and refine the answer with follow-up queries.
‘You can use Bard to boost your productivity, accelerate your ideas and fuel your curiosity. You might ask Bard to give you tips to reach your goal of reading more books this year, explain quantum physics in simple terms or spark your creativity by outlining a blog post,’ Google VP of Product Sissie Hsiao and Google VP of Research Eli Collins wrote in a blog post.
When Google first unveiled Bard last month, there wasn’t much to see other than a lengthy blog post written by Google CEO Sundar Pichai. The model used in Bard is based on Google’s own LaMDA (Language Model for Dialogue Applications) — the company is using a lightweight and optimized version of LaMDA.”
Google went with a “lightweight and optimized version of LaMDA” in order to get it out in the wild without facing a financial strain in terms of computing power needs. One benefit in doing so is being able to build buzz on social media.
An Above Avalon membership is required to continue reading this update. Members can read the full update here. An audio version of this update is available to members who have the podcast add-on attached to their membership. More information about the podcast add-on is found here.
(Members: Daily Updates are always accessible by logging into Slack. If you haven’t logged into Slack before, fill out this form to receive an invite.)
Above Avalon Membership
Payment is processed and secured by Stripe. Apple Pay and other mobile payment options are accepted. Special Inside Orchard bundle pricing is available for Above Avalon members.
More information about Above Avalon membership, including the full list of benefits and privileges, is available here.