Transitions: (in) Artificial Intelligence
is AI here to stay? we think so but now what..
Hi Gatherers! Dede here.
I don’t know about you but this April has felt like a month in decades. It feels like we’re in a transition but to where? we can’t say. Transitions are one of the most interesting things to navigate as well as manage. No one can really predict what is at the end of it or if there’s even an end. The only way is through.
I’m going to try to use the next couple articles in this series to talk about transitions.
I have thoughts about the transitions happening in more spaces than I can count but let me start with technology specifically Artificial Intelligence (AI).
Since the release of ChatGPT in Fall 2022, AI has become for better or worse the new buzzword in tech circles. It was briefly surpassed by poly market and category betting after the recent Nov 2024 elections in the United States, but AI has retained its crown consistently for the last couple years. There is a reason ChatGPT (or just GPT as I would refer to it throughout) and other AI chat apps have remained the top free app downloaded in the United States on the AppStore (iPhone).
What’s the transition you might ask, if these apps are so popular already? Well a couple recent developments:
“Reciprocal” tariffs were recently introduced and then paused by the United States on nations around the world. There’s been a broad debate that AI was deployed to develop the formula used to determine the scope and % of these tariffs. These reports have not been confirmed by the White House but it’s interesting that the report seems plausible.
Voice is having a moment as the method of communication with AI. ChatGPT launched initially as a text format based interface so you type in your prompts and get detailed responses. Over time, OpenAI (creators of ChatGPT) started adding voice features as did other platforms like Gemini by Google and Claude. Why is this a transition? These voice recognition features in AI tools are becoming even more powerful. I’ve tried them and they work too well if that’s even a good qualifier.
AI tools normally have two voice modes: speech-to-speech and speech-to-text. With speech-to-speech, you talk to the AI and it responds back in voice. You can select the voice mode you want it to respond back in, this is where OpenAI ran into trouble and it’s currently being sued by Scarlett Johansson. They tried to use her voice from the movie “Her” but she didn’t grant permission. They used a similar voice anyway.
I mostly work via speech-to-text, so I talk and it transcribes and then helps answer what was asked. You’re basically reading your prompts but in more free flowing speech. Trust me, it works magically. I knew it was the real deal when GPT’s basic voice mode catches small intonations in my accent that Siri in its current form even misses.
Voice stands out as a transition because it tapped into the content creator/influencer kick in this generation where everyone wants to be one. Now you can test out how funny you can be via a non-judgmental AI. The number of TikTok’s and reels I’ve seen with this voice mode continues to be anecdotally high. Maybe it’s my research skewing my viewing algorithm but maybe not.AI in law and policy enforcement. This will probably continue to be under reported! However, after reading a couple articles about the new age defense contractors Palantir, OpenAI, Anduril, etc. forming a collective; there might be smoke here.
“the goal is to deploy these trained models on national security systems.”
It seems AI or a similar capability is being used to identify what international student visas to revoke in the US, or what foreign person to deport or how to parse social media platforms at scale to spot potential threats (this has already been done in the past I bet but it feels more salient now).
Memory, memory, memory! OpenAI recently announced contextual long-term memory for ChatGPT and it is one of the weirdest and dare I say also most important transitions in this list.
With memory, GPT moves from being just a search tool to a true companion /co-assistant. It can piece together your concerns, fears and dreams or anything else you share with it and it can use this info to inform how to answer your future prompts.
I’ve shared some details with it and I’m surprised what it now remembers about me. Probably not far-fetched to believe there’s a future where it calls me by my name but also a future where another app can use this memory API (if it becomes that) to help me decide what to buy, eat or wear. Wow, that’s frightening but a little like that meme, it’s also interesting.
Remember this statement: “ChatGPT will now recall information in two ways — using the “saved memories” that users have manually asked it to remember, and “reference chat history,” which are “insights ChatGPT gathers from past chats to improve future ones,” according to OpenAI.”Artificial intelligence — Everything Everywhere all at once. When Apple finally announced its AI suite of tools for the iPhone and other devices, it aptly called it Apple intelligence for the rest of us! Though it hasn’t shipped all of its promises in the iPhone 16, other AI tools continue to climb. Claude/Anthropic has launched Claude Code and model context protocol (MCP), OpenAI released the last non-reasoning model in GPT 4.5 and OpenAI.fm, Google continues to fight back for the lead with incredible announcements for developers (agent-to-agent developer kit) and for all users (Gemini 2.5, one of the best models out there. Period), Microsoft released a reimagined Copilot just in time for the shipment of copilot+ PCs. On and on the AI train goes. Like the starting age of the internet, it is quickly the knows and the know-nots. Of course there are plenty of great posts (especially on substack) that delve deep into the various tech releases but slowly a chasm is forming. On the one side, AI companies and the developers they mainly focus their designs for. On the other side, the rest of us. Some of us stay up to date but I bet you haven’t used all these great AI tools like you have at least tried GPT.
For instance, Manus (a new AI agent) launched to much fanfare earlier this year and I quickly signed up for the wait list. 4-6 weeks later and I still haven’t created my account even though I got off the waitlist almost a couple days later.
I have tried vibe coding with lovable to build a bot that will help my parents get an appointment date for one of their engagements but I quickly gave up due to auth issues and never went back. On and on, stories like mine litter the AI rising. Stories of some success, a ton of luck, too many tools to remember and so much happening all at once.
What do we do? It is my position that transitions like grief also has its stages. Unlike grief though, It’s in no particular order.
Acknowledgement: Make peace with transitions happening. You can do little to change it so the sooner you do so the better.
Decision: Answer your what/where. What do you hope to achieve with each transition? It can be rooted in anything - reality or not. Just put a boundary around the what. Then answer the where question. Where do you expect the transition to lead to? Answer the where within the boundaries of your what.
Planning: Structure the how. How will you get to your what/where in the transitions, from where you are today?
Move: In this case, Nike says it best “just do it”. Even if you get to this stage and your what/where makes no sense. Or the plans just seem shallow. It doesn’t matter because the worst place you can be in transitions is a place of constant. This is where atrophy happens and if you’re constant you risk falling into the chasm that happens with transitions since you’re neither here nor there.
Okay what do I mean? Let’s put this together.
Acknowledge this – AI is here to stay. I know there are broader questions on regulations, when will AGI happen, data ownership etc. these are great topics and I hope to write more about them. But for a 1st step, acknowledge that AI is happening.
If this transition to an AI world is happening, what/where do you want to be in that future? Does your current job or circumstance fit that? I can’t say but you can. Again it doesn’t have to be in reality, frame what you want. Once you decide the ‘what’ and ‘where’, formulate how you get there.
Let’s say you want to be a voice artist making anime short films that you produce and share with your followers on a media platform of your choice. If AI is here to stay as we’ve agreed then your how can be (1) Get familiar with voice AI (2) Learn how to copyright your voice so you own it (3) If the copyright doesn’t exist, get law involved etc (4) Learn how to copyright your prompts and so on.. I can go on but you get the gist. Pick what (#) you start with and start making progress on it today. Get an account on a voice AI platform like openai.fm. Play around in their toolbox, gauge what works and what doesn’t. Then on to the next step.
Okay to wrap up because I enjoyed writing about this more than I expected. Transitions are here especially in technology via Artificial intelligence. The world is not waiting and neither should your dreams. Go make something happen.
Happy new week! Let’s go
Cheers
Dede