The “open source Ai” fallacy is expanding with mentions from CEOs at major companies on a “global stage” at conferences. There’s now cable coverage, too. The general public is likely becoming aware of the phrase. It’s both enlightening and concerning.
They actually mean locally-infered Ai. And we’ve discussed the misuse of the phrase in the past. Since then, we’ve seen CEOs mention “open source Ai” in live public announcements:
- AMD CEO, Lisa Su at SXSW in an interview.
- NVidia CEO, Jensen Huang at NVidia’s GTC event.
- Meta CEO, Marc Zuckerberg in Instagram reels.
Local Ai is part of the future. And these companies should promote their hardware, as have we with Intel’s 155H and AMD’s 8700G. This is important hardware.
Cable news has challenged the openness of Ai. These executives and investor’s comments should receive scrutiny. It’s good to see. These things matter.
This is part of the challenge of journalism. And when “everyone is doing it,” it can’t be personal. Language is based partly on trends. It’s somewhat fluid, but we believe differentiation is necessary here.
Local Ai is not an unappealing phrase. People love to support local businesses, hire local talent, or visit a local brewery. It’s a meaningful justification to buy new hardware and infer locally. Hardware companies should want to break the cycle of hardware stagnation we’ve seen in consumer tech. Consumers deserve local Ai options.
The user typically decides what is popular, but Ai might be different. Companies often adjust their products in response to criticism from users. Yet in fundamental ways, chatbots have been institutionalized by leadership who believes these services are inherently useful. Consultants like McKinsey say it will bring the 4th industrial revolution. Do the workers agree? The useage stats could tell a different story. How will that impact use?
Companies want to discontinue services that aren’t used. We’ve seen this affect services previously beloved, RIP InVision. But could a preceived underutilization of Ai services convince companies to cancel the services or would it inspire corporate leadership to push Ai harder? The answer will be different for every organization.
Workers should be empowered to voice their concerns. When leadership asks workers to use Ai in a way that is infeasible or unhelpful, and it will, workers should feel comfortable to say it. These tools should be judged on their merit rather than hopes and dreams. Even when a consulting firm claims Ai is generally useful, the burden of proof is on these specific Ai companies to make that case for their tool. Prove it is useful and scrutinize or critique the weaknesses.
Disclosure: Devin owns stock in NVDA and META.