Data Centers Bypass the Grid

Awakend.ai tracks where AI value is moving: off trophy models and into the machines and balance sheets that run them.

This week, Bloom Energy’s revenue jumped 138% as AI data centers stopped waiting for the grid, Broadcom was reported to be arranging $50 billion so OpenAI can build its own chips, and TSMC posted a sixth straight record profit.

The longer look, last, is cancer. AI helped build a vaccine from each patient’s own tumor, and in a trial of 1,137 people it cut the risk of melanoma coming back, or of death, by 42%. Whether it helps patients live longer is still unknown.

Melanoma makes up about 1% of skin cancers, but it causes most skin cancer deaths, and this year about 112,000 Americans will be diagnosed with it, with about 8,510 deaths expected. Caught early, its five-year survival rate is above 99%, but once it spreads to distant parts of the body, that rate falls to 36%. Surgery can remove the tumor, though keeping it from coming back is the hard part.

But medical AI is now helping build a vaccine from each patient’s own tumor and cutting the risk of melanoma coming back or of death by 42%.

Why Melanoma Comes Back
When a surgeon removes melanoma, the scans can come back clear, but tiny clusters of cancer cells can stay behind, too small for any scan to find. If they grow, the cancer returns, which doctors call cancer recurrence, and the risk rises with each stage, even after successful surgery.

High-risk patients can also get Keytruda after surgery, an immunotherapy drug that helps stop cancer cells from using a pathway called PD-1 to hide from the immune system. It helps, but in a new trial, only 70.0% of high-risk patients on Keytruda alone were still cancer-free two years later. For the other 30%, the cancer came back, or they died.

Every Cancer Looks Different
To close this gap, the immune system needs to know what it is hunting, and for decades scientists tried to teach it with a cancer vaccine. The vaccines looked strong in the lab but produced “mostly nothing useful in the clinic.” 

The reason was familiarity. Those vaccines trained the immune system to attack markers that healthy cells carry too, and since the body had spent a lifetime learning to leave those markers alone, it held back.

The markers worth chasing are the ones only cancer makes. When cancer cells mutate, they produce abnormal proteins that healthy cells don’t, called neoantigens, and because those mutations are unique to each person, every tumor carries its own fingerprint.

For this approach, every patient needs a vaccine of their own.

A Vaccine Written From One Tumor
Moderna and Merck built one, called Intismeran, for people whose melanoma a surgeon has completely removed. Scientists compare a piece of the patient’s tumor with a sample of their blood to spot what makes the cancer different, and algorithms pick out as many as 34 of those differences, the ones most likely to trigger the immune system. The vaccine carries them into the body as a set of instructions, so the body can make those cancer markers itself and show them to the immune system.

The companies call it an individualized neoantigen therapy, a medicine made for one person’s cancer. 

Think of it as a wanted poster made for one criminal. The vaccine shows the immune system that cancer’s face, and Keytruda takes away its hiding place. No two patients get the same poster.

The Real Test
In a trial of 1,137 people whose high-risk melanoma had been removed, everyone got Keytruda, and two in three also got their personal vaccine, while the rest got a placebo, a stand-in with no vaccine in it.

Two years later, 82.2% of the patients who got the vaccine were still cancer-free, compared with 70.0% of those who didn’t. Overall, adding the vaccine cut the risk of the cancer coming back, or of the patient dying, by 42%.

The companies say it is the first time a treatment like this, built for each patient, has succeeded in a final-stage trial. But the results reached the public earlier than planned. Screenshots of the data leaked on social media two days before the planned presentation, and Moderna’s stock, which had more than tripled since August, fell 17.6% that day. Investors had expected a 40% to 50% cut in risk, and 42% landed near the low end.

The trial can’t yet show whether the vaccine helps people live longer, and its lead researcher says that answer will take more time. Regulators haven’t approved it yet.

Why It Matters: The real story here isn’t melanoma. It’s that cancer medicine is starting to be made for one person at a time. Doctors call this precision medicine, and so far it has mostly meant using tests to pick the drug most likely to help you. This goes further, because the medicine itself is written from one patient’s tumor.

For patients, that changes what comes after surgery. Instead of waiting and hoping the cancer stays gone, they could start a vaccine built for their own cancer. With trials underway in lung, bladder, and kidney cancer and a possible approval in 2027, the next person who has a tumor removed may leave with a medicine made only for them.

Personalized Cancer Vaccines Are on Track to 4x by 2035

Takeaway:
The money is moving toward medicine made for one person. Tumor-specific vaccines already hold 48% of a market growing 18.7% a year, and analysts expect intismeran to bring in $6 billion or more a year in melanoma alone if it’s approved. The trial showed the vaccine can work. The next test is making tens of thousands of individual batches without repeating Provenge.

The Timeline Guide:

Each timeline and graph represents the realistic stage of the covered technology, plotted from concept to scale, capturing where it stands today and when broad deployment is likely.

Hit reply with anything else on your mind; we actually read every note.
See you in the next upload

Note: Images in this newsletter are often AI generated for illustrative purposes only. Market forecasts are hypothetical, inherently uncertain, represent our best-guess estimates only, and should not be considered investment advice.