Cardano founder Charles Hoskinson says the blockchain’s engineering culture has pivoted decisively from the “monolithic purity” of its early years to a multi-team model that—if it works—will finally give the ecosystem the development velocity many stakeholders have long demanded.
Why Cardano’ Engineering Is At An ‘Inflection Point’
In a post on X, Hoskinson reflected on the methodical, research-first mentality that produced Cardano’s Haskell-based node and its peer-reviewed protocol suite, but also, in his words, “deeply alienated many who wanted to contribute, prevented new ideas from entering our ecosystem, and resulted in a very slow delivery of features.” He described the moment as “a very exciting inflection point” marked by the coexistence of “small agile teams like Aiken and Midgard” with the heavyweight formal-methods units at IOG—particularly the group that continues to specify critical components in Agda.
“We are opening up the ecosystem,” Hoskinson wrote, adding that the arrival of alternative clients “will and should challenge the assumptions, designs, and features of prior teams to accommodate new protocols, economic realities, and our progress as a community.” He conceded that the transition would displease those “who will look for easy scapegoats and want a return to days that never actually existed,” but insisted that “the new days will be filled with even greater opportunities and exciting new capabilities.”
Hoskinson’s upbeat framing clashes with a wave of concern triggered by IOG’s termination, on 30 April, of its long-running contracts with Well-Typed and PN-Sol—two specialist vendors that supplied much of the low-level talent for Cardano’s networking, scalability and key-management initiatives. The move surfaced publicly during the Technical Steering Committee (TSC) meeting of 7 May, whose minutes were summarised by community member YUTA-Cardano/CPA.
According to that summary, engineers from the two firms had been working on Leios, a high-throughput execution layer targeting roughly 1,000 transactions per second; on the Log-Structured Merge-tree (LSM) project, which relocates UTxO data from memory to disk; and on KES Agent, an externalised key manager meant to deliver the forward-security promise of the original Preos design.
Duncan Coutts, a veteran Cardano architect now serving on the TSC, warned that “one-third of the networking team may have been lost” and that onboarding replacements “typically take over six months,” raising the spectre of delivery slippage at the very moment when Hoskinson is pressing for acceleration.
Leios Release In 2026
When a user on X asked, “Why were key members contracts terminated?”, Hoskinson’s reply was unflinching: “Because I want Leios in 2026 not 2028 and I value Pragma and different ideas and implementations. No more fucking games or delays. Cardano needs to get to the next level.” The remark confirmed that internal deadlines—not performance issues—lay behind the shake-up, and that IOG is prepared to reshuffle contractors if it believes alternative teams can move faster.
Hoskinson’s call for “no more…delays” crystallises an old tension at the heart of Cardano: the project’s reputation was built on mathematical caution, yet its market relevance depends on proving that scholarly methods can deliver at commercial speed.
At press time, ADA traded at $0.81.
