I agree, agents are the next step in this evolution and for that to be a game changer we need
- Discovery, agents should able to discover new systems and autonomously interact with them
- Better Experience, I think AI experiences are where we were with mobile experiences around 2010, we need to figure out how to create better human-ai experiences. Right now it's probably faster and easier to do a lot of thing yourself than to use an AI assistant/agent except for a few specialised ones that are integrated into existing workflows.
I see a lot of anticipation for GPT-5, I believe what we should be looking forward to is some form of autonomous agency where, it can interact with everything from digital systems to the physical physical world without need for manually integrations as we have today or at least digital systems. Without that a more intelligent GPT will not be a big deal.
While it's pedictable that GPT-5 is going to be better than GPT-4, at the same rate as the previous generation.
I think the problem is less of building products on top of GPT-x, but more of build a general-purpose product that can easily overlap with OpenAIโs mission.
Because @OpenAI is pursuing a broader vision of AGI, which means they won't be able build for specific verticals. Atleast until GPT can autonomously build companies.
For now, while everyone can use @ChatGPTapp, it's too general purpose and detached from users everyday workflow in different verticals.
For example, I use @github co-pilot everyday, although it's basically built on GPT, the fact that it's integrated into my workflow makes it way better than the @ChatGPTapp for coding task.
So I'm betting on products that integrate GPT into everyday workflows, for those products, GPT getting exponentially better is a net positive for them.
@30nemagar I believe the objective for every iteration is basically to get a step closer to AGI, but getting to AGI would require more tan just LLMs, it would require some form of autonomous agency where, it can interact with everything from digital to physical autonomously.
@gregisenberg At early stages, I don't think the last point holds because if responsibilities are adequately shared, everyone would be wearing multiple hats, all of which would be stressful regardless of whether you're the CEO or not
๐ Just a few days ago, I launched Tethered v2 in beta! It was accessible to anyone who visited our page, and to my amazement, within less than 24 hours without any outreach,
๐ก Taking inspiration from their feedback, I started working on Tethered v2, incorporating their feedback. Additionally, I worked tirelessly to find ways to significantly reduce costs, as the freemium version was already incurring expenses of around $700 within the first month.
@AdoraNwodo Iโve been waiting for someone to point this out for a while now.
Iโve seen a lot of these over the years and itโs just disheartening because it just breeds a new generation of shallow engineers.
You will fully grasp the depth of the decay when you try to hire.
@Egi_nupe_ Seriously painful, considering how worry about everyone else in our lives before ourselves.
I for one, I donโt even spend up to 20% of my income on myself.
When I look at my expenses monthly, they are mostly money sent to other people in my life for different things.