OpenAI Developer Day – Tanks on Grass

Sam parks his tanks on Sundar and Mark’s lawn.

  • OpenAI has launched a series of new products that aim to capitalise on its renewed momentum in the enterprise, wipe out Meta and Google’s existing ecosystems and make OpenAI the most attractive place for users to spend time and to buy and sell AI services.
  • OpenAI has also decided to postpone its IPO still further until such time that it has greater certainty about how its products will behave when let loose amongst the general public.
  • In just under 60 minutes, Sam Altman launched and demonstrated a large range of products of which the most important were:
    • First, Dots: which are agents personalised to the user that are effectively an executive assistant for the user’s life and the user’s job.
    • This is a virtual copy and paste of what Meta has done with Muse and sets the tone for the next level of competition in the AI services industry going into 2027.
    • However, just like everyone else, users are going to be wary of handing over their email accounts and credit cards to these agents, which is what they need to do if they want the agent to be useful.
    • The problem is that agents remain pretty unreliable, and so take-up and adoption is going to take some time.
    • Sam has once again parked his tanks on Sundar and Mark’s lawn, but the castle is very well fortified and continues to have billions of users using its services every single day.  
    • Second, models and APIs: where OpenAI is refining its product suite to cater to what developers have asked for.
    • These include a new model GPT6.1 Sol, which is almost as good as Astra but costs 1/5th of the amount.
    • OpenAI also launched Ultrafast, where the user can pay extra to get their tokens more quickly, which is important in some use cases.
    • For example, the top tier offers tokens 8x faster at 6x the normal price.
    • It also launched the Decisions API, which is basically a copy of the Jev product that has gained quite a lot of attention recently.
    • This is an API that helps software make decisions that it can’t be done in software, such as deciding if a customer email sounds annoyed, allowing the software to flag it.
    • Third, App Store 3.0: was launched for the 3rd time and this time it is called OpenAI Marketplace.
    •  It is a place where developers can post their AI services and users can discover and buy them, which is the backbone of any digital ecosystem, and explains is why OpenAI will keep on trying until it gets it right or loses the war to Google, Meta or Apple.
  • OpenAI has also decided to postpone its IPO until such time that it can “make confident safety decisions”, which is an attempt to close down the endless stream of negative press attention that the company has been receiving.
  • This is a sensible approach that could be made a lot better if the company were much clearer about the areas where it is concerned about the behaviour of its products.
  • However, it does mean that the company needs to raise more money, which is reportedly being offered at a valuation of $1.4tn.
  • The net result is that this developer day is exactly the same as the other ones that have gone before it, except that this time, it is focused on both the enterprise and the consumer rather than just the consumer.
  • Once again, it is abundantly clear that OpenAI knows that it has to own the touchpoint with the user, which is increasingly looking like an agent of some kind or face turning into a commodity.
  • At a mooted valuation of $1.4tn, it cannot afford to get this wrong, and it will have to overcome at least one of Google or Meta, who won’t take this lying down.
  • Meta has already come out swinging, but Google is still in the dressing room contemplating its navel.
  • I have a position in Meta in anticipation of the new enterprise business that it launched on Monday, but if renewed AI momentum takes the shares to my target, I won’t be complaining.

RICHARD WINDSOR

Richard is founder, owner of research company, Radio Free Mobile. He has 16 years of experience working in sell side equity research. During his 11 year tenure at Nomura Securities, he focused on the equity coverage of the Global Technology sector.

Leave a Comment