Daily Brief — April 2: Lightning Splices, ALS Breakthrough, Grid Record

The pattern underneath this morning is simple: the technology meant to solve our hardest problems is moving faster than the systems designed to deploy it. Lightning gets splicing. Brain-computer interfaces decode speech from the locked-in. Solar and batteries are about to add a record 86 gigawatts to the American grid. And then the grid itself becomes the constraint.

CORE LIGHTNING v26.04rc2 SHIPS WITH CROSS-CHANNEL SPLICING

The Core Lightning team released v26.04 Release Candidate 2 today, and this one matters because splicing technology is no longer an advanced feature — it is becoming the ordinary way people move money between channels. The headline additions: `splicein` and `spliceout` commands for adding and removing funds from channels, plus a new cross-splice capability that routes funds between two channels in a single operation.

Payment reliability improved through parallel pathfinding and bug fixes in the `askrene` module. Payers can now add notes when paying via `xpay`. The release includes stricter conformance to BOLT spec for splice commitments.

Why this matters: splicing eliminates the need for multiple scattered channels per user, which scatters liquidity. Wallet providers can now abstract away the distinction between on-chain bitcoin and Lightning balances through unified wallets — the distinction becomes largely irrelevant to the user. The Bitcoin network is running at ideal conditions for on-chain splice transactions: 986.3 EH/s hashrate, fees at 2 sat/vB fastest, 1 sat/vB for 30-min or 1-hr confirmation.

The release candidate came from @rustyrussell, @ShahanaFarooqui, @sangbida, @endothermicdev, @cdecker, @nepet, @Lagrang3, and @niftynei.

BCI BREAKTHROUGH: SPEECH DECODED FROM LOCKED-IN ALS PATIENT

A study published this week in *Cell Reports* and indexed in PubMed demonstrates the first successful real-time speech decoding via intracortical brain-computer interface in a person with long-standing anarthria and locked-in syndrome — meaning someone who has completely lost the ability to produce any speech movement.

The participant had not spoken in over two years. Previous iBCI demonstrations relied on patients who could still attempt to move speech articulators or vocalize. This is the most severe form of communication loss.

The researchers decoded phonemes, words, and higher-order language units well above chance. Sentence decoding accuracy was below that seen in participants with dysarthria — but the team hypothesizes that higher-level encoding of intended language output at the sentence level may be preserved even in long-standing anarthria, and proposes ensemble model sentence decoding to improve accuracy.

This work sits at the intersection of machine learning and robotics — the exact domain *Nature Electronics* named as its technology of the year for 2026.

U.S. GRID ADDS RECORD 86 GW IN 2026 AS CCS GETS SHARP CRITIQUE

Two developments define today's energy landscape. First, the EIA data: U.S. power plant developers plan to add 86 gigawatts of new utility-scale generating capacity in 2026 — a record if realized. Solar comprises 51% of planned additions, battery storage 28%, wind 14%.

Developers plan to add 43.4 GW of new utility-scale solar (a 60% increase from last year), with more than half going to four states: Texas (40%), Arizona (6%), California (6%), Michigan (5%). On storage, developers plan 24 GW of utility-scale battery additions in 2026, compared with 15 GW in 2025. U.S. battery storage capacity has grown exponentially over the last five years with more than 40 GW added during that period.

Second, the constraint becoming visible: a *Nature* letter published this week by Arendt challenges the fundamental accounting of carbon capture and storage. After fifty years of commercialization, global CCS capacity has reached 0.09% of global emissions. Even if installation rates immediately expand tenfold, this will make no quantitatively important contribution to climate mitigation by 2050.

The real story is not CCS. It is solar and batteries moving at speed. The structural bottleneck emerging is grid capacity itself. Over 2,500 GW of renewable, storage, and large-load projects are stalled in grid queues worldwide. Meeting electricity demand through 2030 will require annual grid investment to increase by approximately 50% from today's USD 400 billion.

The technology is ready. The grid is not.

BY THE NUMBERS

86 GW of new U.S. utility-scale generating capacity planned for 2026 — a record, with solar and batteries comprising 79% of the total.

24 GW of utility-scale battery storage planned for 2026, compared with 15 GW in 2025 — a 60% year-over-year increase.

2,500 GW of renewable, storage, and large-load projects currently stalled in global grid connection queues — the bottleneck that will define deployment through 2030.

0.09% of global emissions — the current share covered by carbon capture and storage after fifty years of commercialization.

SIGNALS TO MONITOR

Core Lightning v26.04 final release (expected within days, following this RC2 drop). Watch whether adoption of splicing by wallet providers accelerates to the extent the UX distinction between on-chain and Lightning balances meaningfully disappears for users. This is the inflection point between Lightning as a technical solution and Lightning as a platform.

U.S. grid capacity additions through Q2 2026 (next update mid-year). Watch whether the record 86 GW figure is realized, and whether queue lengths for new projects shrink or continue growing. If queues grow despite record additions, the grid bottleneck is structural and investment must increase faster than current projections suggest.