Yesterday we spent the day in Madrid meeting with Sam Altman and the OpenAI team as part of their world tour. We were 30 invited developers that sat down and had a one-hour conversation about what OpenAI are working on, some of the improvements that we would like to see, what is on the roadmap and how they're dealing with EU-specific issues.

No surprise, Sam Altman is a superstar. It was like seeing a rock star in action. Not during the developer meetup, but definitely afterwards as we went to the bigger conference hall with 650 people. And the meet and greet afterwards.

Just before coming to the meeting they just met with the Spanish government and they are keeping a close eye on what's happening with the European AI bill that's being worked on. They are hopeful that Europe will not get cut off from generative models.

In terms of their product roadmap, the number one thing on their mind is speed and latency. That's also their number one request, both from developers and from ourselves.

One of the most interesting and relevant aspects of the roadmap for us is the Sam mentioned that within this year they will be launching a voice interface, which is surprisingly fast and allows you to interact with GPT using voice like a real conversation. This will be available using the API as well. 

We also connected with Mo, one of the research scientists at OpenAI, who promised to give us access to the 32K model, even though that will not be production ready. 

We had a couple of people getting very excited, and especially one of the attendees, another fellow entrepreneur, who has a mother with Alzheimers. When he heard that we're building a digital twin that possibly could help with the communication challenges, he was overly excited. This made us discover another very important use case to deal with family members that need support and help.

We also discussed the opportunity of using channels. We've already seen people building stuff using Telegram, and during the meetup I met another developer that has been building a WhatsApp version of ChatGPT. This made me think about the opportunity for us to use WhatsApp as an access point to what we're building - more to come on that soon.

Altogether it was a high energy trip back and forth to Madrid over the day, and super exciting to be part of this small but very interesting community of 30 AI driven developers. 

Meeting Sam Altman at OpenAi Developer Round Table May 22, 2023 in Madrid

Haven’t we all wished that we could clone ourselves? There are many reasons we would want to do that:

 A digital twin could provide a unique form of companionship, as it would share the same values, interests, and experiences as the individual.

 Having a digital twin will allow people to gain insights into their own behavior and thought processes, and facilitating self-improvement.

 A digital twin could help an individual accomplish tasks more efficiently by dividing workloads and responsibilities between the two identical beings.

 Cloning might be viewed as a way to replace a lost loved one or to recreate an individual who has passed away.

While we believe that these are super interesting problems to solve for, and we will eventually get there, we are initially focused on one specific problem:

How could I help more people by scaling my knowledge and access, infinitely?

 How could I do it without getting burned out? And how can I do it in a way that allows me to monetise my knowledge?

Providing personalised, one-to-one interactions at scale remains unsolved

Being able to provide unique, personalised value to a large group of people sits somewhere between incredibly time consuming and impossible. In fact 13% of all full-time creators say they are extremely/constantly stressed. (

Influencers and thought leaders simply do not have enough time to cater in an individual and personalised way to their entire community. Most people have too busy schedules to be able to interact in an unique way with their fans, employees or sometimes even the people closest to them.

Instead they use the one to many approach - a generic, one-size-fits-all experience, which does not resonate or help each person in their community in the most effective way.

This lack of personalisation limits the potential for deep connections and meaningful engagement, as well as the impact that influencers and businesses can have on their followers or customers.

From a revenue perspective there is also a missed opportunity to charge for a more personal experience.

Simply put - too little time, too many people to help… painful!

“I’ve been very very very close with my audience, but when you have hundreds of millions of views every single month, it’s just not humanly possible to speak to every single viewer,” says Caryn Marjorie, 

,, who posts over 250 pieces of content to Snapchat every day. “And that’s where I was like, ‘You know what: CarynAI is gonna come and fill that gap.’” And she believes the company has the potential to “cure loneliness.”

Access is kinda broken - we all want unique and personalised access to the people we care about

The flip side of this is that we all crave tailored experiences and authentic interactions, to learn and grow from the people we admire and to have unlimited access - if we could. Just imagine what it would feel like if you had infinite access to your idol, maybe it’s an artist that you love, or a business man, a life-coach or a doctor. Imagine that you could access everything they have said, every advice they could provide, through a natural conversation, and get instant answers tailored to you, instead of having to search through articles, video clips or course material.

Twise makes it easy and enjoyable for anyone to create a digital version of themselves. We call them digital twins.

A digital twin speaks like you, acts like you, and (eventually) looks like you. It uses your actual voice, through voice cloning, your tone and style, and has your personality, your experience and knowledge.

Each twin has a short term memory, it remembers every single interaction and conversation, it has long term memory that get’s better over time and can store an (practically) infinite amount of knowledge and stories. It also has a dynamic personality that evolves over time which becomes more refined as the creator interacts and “trains” the twin.

As of May 2023 we have already built all of the above. Except for the video/visual cloning which is still an opportunity to be solves as real-time video generation is not yet possible (but give it a few months).

We still have a way to go until we have a proper scalable solution, our first developer just joined 10 days ago as we write this, although we are making rapid progress.

An introduction to Twise.ai

 there will be millions of people that will want to have a digital version of themselves. Some people will use it for personal reasons to get to know themselves better, maybe use it as a diary, as a speaking partner, as a problem solver, or as an assistant that will do things for them but will be mainly private for personal use.

 there will be hundreds of thousands of people that will want to take their digital version of themselves, their twin, and make it available so that others can interact with it. And some of those people will want to charge for that. Thought leaders, opinion leaders, influencers, educators, etc.

 we are the right team and at the right moment in time to do this. We are capable of taking this to market and make something unique out of it, because of our experience in scaling technology businesses (Rene helped grow Mural from $7m to $120m), developing AI powered software (we’ve been building side hustle products since early 2022 using the GPT API) and impacting millions of people through creative communication campaigns (Jab has created campaigns for Nike, Ikea and Samsung that have reached millions of people).

Wisepunk

Three things we believe

More Stories

Meeting Sam Altman at OpenAi Developer Round Table May 22, 2023 in Madrid

An introduction to Twise.ai