In our own company, our tooling has mostly sucked ( after a period of being decent ). Any suggestion of better ( in this case more expensive models ) is rebuffed as both cost saving measure and efficiency. We are basically told to work not only with corporate handicapped versions of those models, but also ones, where company can save a buck and pretend it is going well.

Its not gonna end well, but, because it was clearly not a technical conversation, I smiled and nodded.

edit: In case you are wondering why it is relevant. OpenAI can do all sorts of fun things, but it is largely pointless if corps are not using it to its potential. And they don't.

The goal over the next few years is to reduce the "high level" human team member count as much as possible as more work gets assigned to agents.

It will start out as really helpful agents but very quickly lead to the elimination of roles.

Which I think overall automation should be a good thing. But it makes it more urgent to correct the structural flaws in society, starting with classism, Social Darwinism, etc.

Does this mean custom gpts are a business only feature or will they have Workspace agents for Consumers?