Products like Gemini by Google indicates the potential for larger companies to respond to the AI trend and respond. A well financed Autodesk seems to be thinking on this level (see their website) I would find it difficult to say that Revit will be obsolete. The practical application within BIM and architectural design tools like Revit involves complex integration challenges, industry acceptance, regulatory compliance, and substantial training and adaptation periods. This comprehensive approach of BIM means outcome driven processes based on data will require signifcant coordination of that data. AI will always be detached from reality, and until we can 'trust' AI we will always have some form of human interaction, need and oversight. The aforementioned gap is a significant gap that will pose challenges for decades to come, just as the invention of electricity caused concern people (they thought it would kill birds). If AI is posited as a tool, it can certainly help Revit but not replace it. Its like saying the 'massing' tools as a concept can replace Revit. If AI is posited as something more, then its got a large hurdle to overcome.
@@DD-eq2bl Hmm good question. I doubt AI will serve as a 'replacement', but rather more in terms of integration. I anticipate a (long distance) future where we can type 'car' into a part of Rhino, or an extension/plugin of Rhino, and it can model a 3D car. There AI's that currently do this. Yet, the trick will be complex prompts. For a prompt like this 'Design a 1200x800 Cottage window, doubled glazed, using Truecore Alspec Profiles', it will require signicant understanding fo AI on what all these each part is, to the point you will likely need to feed in the data: shapes, sizes, and the profile, which then is leaning more into parametric design. Specificity, accuracy, data to train on, and integration are all hurdles that isolated GPT text based AI models dont need to worry about. Im not stressing at all really. Im looking forward to being impressed, but in my line of work AI has not touched at the level of specificity and compliance that is required for AI to be valuable.
It seems to me that this guy is full of standard lines like "You can't do the same thing and expect different results", spreading fear of missing the AI train, and generally full of not-good stuff
Do you remember when 3D printing technologies experienced a revival (although they had already existed since the 1980s) and suddenly there was talk of everything being printable, with other manufacturing technologies becoming obsolete? Also, with Blockchain technologies, hundreds (thousands!) of new companies and apps emerged promising a new "salvation," and now that's what we see. With AI, it will likely be the same. It's just another tool, albeit with potentially greater capabilities in this case, but ultimately it won't completely replace anything. By the way, I'm a power user of these three technologies.
