Does Larry Goldberg actually target a $1,500 Tesla price by 2027?
No. He says the $1,500 figure was Randy Kirk's suggestion on another show, and his own response was only that it is not impossible.
Goldberg stresses he was reluctant to give any price target at all and views 2026 as the year the strategy lands, with the stock repricing expected during 2027 rather than tied to a specific target.
He believes multiple catalysts reaching maturity in 2027, led by robotaxi but including Cybercab, Semi, and Optimus use-case trials, will give Wall Street the information needed for a significant rerating of the stock.
When does Goldberg expect unsupervised FSD and what does it mean?
He expects unsupervised FSD around late 2027, likely in the United States first, after regulators set a standard before the next election. In his definition, the driver need not be in control or paying attention, but may still need to sit in the driver's seat as a precaution.
He argues current geofencing is a containment and management strategy, and that the shackles come off during 2027; if that slips to 2028, the timeline moves back.
His confidence rests partly on his own experience driving FSD for a year and a half without interventions beyond personal preferences, while acknowledging he is an end of one and only Tesla has the fleet-wide data to know how close the system is.
Can the robotaxi business really reach a dollar-per-mile price point by 2027?
Yes, he believes sub-dollar-per-mile rides are achievable in the Cybercab by next year, and notes Tesla is already charging under $2 per mile with only a handful of cars. Family rides and people giving up cars become viable closer to 70 cents per mile.
Roughly 70% of rides involve a single occupant, which the dollar-per-mile convenience and commute service can easily replace, and he argues owners forget a personal car sits unused all day when comparing costs.
He sees geofencing disappearing by mid-2027 and notes the biggest growth potential lies outside major cities, where three-quarters of US miles are traveled and where robotaxi would replace human-driven cars, a far larger market than today's ride share.
Why does cutting ride prices to a dollar per mile not conflict with showing Wall Street earnings growth?
Goldberg dismisses the tension, arguing Tesla chases its mission as fast as it can rather than trying to convince investors, and that reaching the real market sooner matters more than today's share price.
He says he has no sentiment about retail shareholders pulling their hair out and is looking purely at the value of the company over time.
Do the Semi and energy businesses matter for the 2027 thesis?
Yes. Goldberg argues these businesses already cover Tesla's expanded overhead, so every dollar they earn flows straight to profits, and a 50,000-truck annual Semi run rate would pay substantial overhead and add to gross margin.
He expects Elon Musk to cut overhead with an axe between now and the end of next year, keeping the company lean without affecting growth or innovation, and sees robotruck adoption as not material to the short-term narrative.
Will Optimus robots be sold to customers by 2027?
Not at scale. He expects proofs of concept with perhaps five to ten thousand robots spread across five or ten customers, with volume deliveries going to Tesla itself and possibly SpaceX, and customer use-case trials serving as the real catalyst.
He anticipates bake-offs at many companies testing multiple robot vendors rather than single-supplier commitments, and agrees the use-case results will add foundational layers to the stock's repricing.
He cautions that embodied AI is dramatically harder than pure software AI, so scaling robots to tens of thousands will require far more work than the rapid scaling seen with models like GPT-2.
Why hasn't Tesla launched AI data center leasing revenue?
Goldberg says Tesla simply lacks compute headroom: its data centers are fully occupied training Optimus and FSD, and the delayed release of FSD V15, a 10x parameter model, is an indicator of that shortage.
The company needs AI5 chips in its data centers as soon as possible, and building general-compute data centers for leasing would have forced Tesla to raise capital earlier while it is already stretched across a Texas solar factory and three growing Austin factories.
How does Goldberg answer the objection that a 300-to-1 P/E already prices in the robotaxi rollout?
He points out that Tesla reached roughly $450 to $500 as a pure car company at a 20x multiple, and argues car profits, FSD profits, and burgeoning new businesses will expand significantly with no impediment to a major rise in the stock.
The market is already fairly convinced Tesla is on track for something seriously good and convinced about Elon, though political baggage weighs on sentiment; as these stories deliver next year, he expects a significant expansion of confidence and a rewriting of the stock.
What assumptions and caveats sit around the whole 2027 case?
The analysis excludes any SpaceX-Tesla merger, treats a $1,500 figure as merely not impossible rather than a target, and acknowledges that if unsupervised FSD approval slips to 2028 the timeline moves back. Tesla's own chip production was also delayed, forcing a shift to Nvidia.
The pivot from car company to embodied AI company was made without announcing it or telling management, capital requirements are exploding with tens of billions potentially needed from capital markets, and that spending has been a drag on the stock while the market is not yet aligned with the strategy.