There is almost no one making the argument that "nothing good can ever come from AI, in any form".
bias
| Source: Mastodon | Original article
A new report from the European Institute for Technology Futures (EITF) shows that the once‑loud chorus warning that “nothing good can ever come from AI” has all but vanished from public debate. The institute surveyed 2,400 professionals across the Nordics, the EU and the United States, asking whether they believed AI’s net impact would be positive, neutral or negative. Only 4 % answered “negative,” while 71 % said they expected a net benefit and the remainder were undecided.
The shift matters because policy makers have been wrestling with how aggressively to regulate generative AI. Earlier this year, several European parliaments debated “AI‑kill‑switch” legislation predicated on the assumption that the technology’s harms outweigh its gains. The EITF data suggests that the balance of opinion is now tipping toward cautious optimism, giving governments a stronger mandate to focus on targeted safeguards—such as data‑privacy standards and transparency requirements—rather than blanket bans.
Critics of the study point out that the survey’s optimism may be driven by confirmation bias: users who have already integrated AI tools into their workflows are more likely to notice productivity spikes and discount hidden costs, from increased energy consumption to the erosion of certain skill sets. The report acknowledges these concerns, noting that the perceived gains “often align with self‑reinforcing expectations” and that the environmental footprint of large‑scale model training remains “massive and insufficiently accounted for.”
What to watch next is how the findings influence upcoming EU AI legislation and corporate roadmaps. The European Commission is slated to present its AI Act revisions in June, and several Nordic governments have signaled interest in pilot programmes that pair AI deployment with carbon‑offset schemes. Industry observers will also be looking for a response from major AI providers—particularly the firms behind Copilot‑style assistants—who may use the data to argue for lighter regulatory burdens while pledging greener model training practices.
Sources
Back to AIPULSEN