“@GailAlfarATX @Tesla @aelluswamy Beta expanding to safety scores above 80 after 10.69.2.1 goes out”
The tweet archive.
15 years of Elon, fully searchable. The production archive uses Supabase as the source of truth, with 94,952 indexed tweets available in development as a full-archive fallback and a curated annotation layer for context, theory, and how major claims aged.
“FSD Beta 10.69.2 rolling out now. Please note that top priority is safety, so expect it to be overly cautious, especially around pedestrians.”
“@Erdayastronaut @PPathole @planet4589 @NASASpaceflight Base of the vehicle seems ok by flashlight. I was just out there about an hour ago. We shut down the pad for the night for safety. Will know more in the morning.”
“@DBurkland Team is working the weekend to get 10.12 in limited release tomorrow. Then we evaluate, do a point update, and broaden the release. 10.12.2 is probably where we can expand to a safety score of 95+.”
“@TeslaOwnersEBay @GailAlfarATX @Tesla @EvaFoxU @Proxima06_Aiai @DBurkland @lewins_ian @Lstad78 @1stMarsColonist @JosephVVallace @dubbleotrippleo No, I meant wide release to high safety score beta participants. Going to all beta participants is expected later this year.”
“@DriveTeslaca Yes, but we take a very cautious approach to safety. As we confirm functionality on Canadian roads, the beta will expand.”
“@jchybow If this version performs well, we can probably lower min safety score to 95”
“@DriveTeslaca Depends on what issues we see that are unique to Canada. If things are looking good, then we can lower the safety score threshold.”
“@NoahWebb_ @WholeMarsBlog @tkrisher Will begin limited rollout to Canada this month. Safety is paramount, so we have to confirm no significant issues, given slightly different road rules.”
“@WholeMarsBlog @tkrisher He’s actually a lobbyist, not a journalist. There are many who pose as the latter while behaving like the former. No integrity. Indeed, there were no safety issues. The car simply slowed to ~2 mph & continued forward if clear view with no cars or pedestrians.”
“@24_7TeslaNews @WholeMarsBlog We are rolling 10.4 out slowly, as there are many subtle hardware differences in the owner fleet. If it continues to look good over the next few days, we may start rolling out to 98 safety scores.”
“@jgrano305 @WholeMarsBlog Tentative plan is 98 & above starts uploading Friday afternoon next week. If we see any concerns, uploads will pause while we investigate, so might take a few days before everyone with 98 safety gets beta 10.4. 10.4 improves left turns across fast traffic & stopping for gates.”
“Beta 10.3 releasing Friday next week to all cars with 99/100 safety score”
“Beta 10.2 now rolling out to cars with 100/100 safety score over 100 miles”
“@johnkrausphotos Definitely further refinements coming to (early beta) safety test score. It will be refined continuously until it is an extremely good predictor of crash probability. Exciting actuarial problem!”
“FSD Beta 10.2 rolls out Friday midnight to ~1000 owners with perfect 100/100 safety scores. Rollouts will hold for several days after that to see how it goes. If that looks good, beta will gradually begin rolling out to 99 scores & below.”
“Wow, lot of interest in FSD beta! Plan is to roll out version 10.2 midnight Friday, then on-ramp ~1000 owners/day, prioritized by safety rating.”
“@teslaownersSV Maneuverability is important for safety”
“@WholeMarsBlog Expansion will be gradual, based on ensuring maximum safety”
“@mrgreyev @DirtyTesla @Tesla When we need strange training examples, eg half a car on highway, Tesla sim delivers. Also, accidents on Autopilot are becoming rarer, so we need more close call examples. For safety tests, Autopilot sees dummy far away. Emergency braking only needed for distracted drivers.”
“@IIHS_autosafety Top Safety Pick+ from IIHS! That’s the highest overall rating, but we think we can get max score for all subcategories/individual tests too. Retesting in a month or so with improved software. The improved software will then be uploaded to all cars with FSD computers.”
“@marcodzo @Tesla Sometimes people think it is performance, but safety is always the primary design requirement for a Tesla”
“@PPathole @DELTA_V @spacex360 After several successful launches, land overflight earlier in trajectory passes E-sub-c safety threshold. That said, Starship will also launch from Cape long-term.”
“@garyblack00 We had to focus on removing radar & confirming safety. That release goes out next week to US production. Then a week or two to polish pure vision FSD & v9 beta will release. Difference between v8 & v9 is gigantic.”
“@teslaownersSV @TeslaNY @Tesla Gating factor is achieving & proving higher safety with pure vision than with vision+radar. We are almost there. FSD Beta V9.0 will blow your mind.”
“@hamoon__ @neuralink Neuralink is working super hard to ensure implant safety & is in close communication with the FDA. If things go well, we might be able to do initial human trials later this year.”
“@comma_ai Tesla Full Self-Driving will work at a safety level well above that of the average driver this year, of that I am confident. Can’t speak for regulators though.”
“@OwenSparks_ @flcnhvy @Tesla Safety is our primary design goal”
“Dragon can also hover & land propulsively, but we use the parachutes for maximum safety https://t.co/Mr7VFIQwWf https://m.youtube.com/watch?v=07Pm8ZY0XJI”
“Thrusters are integrated with spacecraft, enabling abort capability all the way to abort for increased safety”
“@Teslarati Dependent on proving high safety”
“@jwangARK Tesla FSD computer’s dual SoCs function like twin engines on planes — they each run different neural nets, so we do get full use of 144 TOPS, but there are enough nets running on each to allow the car to drive to safety if one SoC (or engine in this analogy) fails.”
“@ValaAfshar @Tesla Having a low center of gravity makes a world of difference to safety”
“@Erdayastronaut @SpaceX All correct. Parachutes were originally the backup landing system, with SuperDraco thrusters as primary. Difficulty of proving thruster landing safety *and* architecture being suboptimal for moon/Mars caused us to change focus to parachutes.”
“@flcnhvy No, best to have people seated for safety & speed of travel”
“Tesla factory workplace safety is 5% better than industry average for manufacturers & improving fast https://t.co/axmInA9fZb”
“@flcnhvy Tesla makes the safest cars in the world & our factory injury rate is better than industry avg & tracking to be better than any auto plant in North America. We care deeply about safety at a fundamental level.”
“@PPathole Pretty much. Being better than human drivers on average is not a super high standard. Autonomy will probably be held to a 300% to 1000% higher safety requirement long-term.”
“@TeslaGong @_KarenHao @OpenAI I have no control & only very limited insight into OpenAI. Confidence in Dario for safety is not high.”
“@CathieDWood @sbarnettARK @jwangARK @ARKInvest First, we need to make it super safe & easy to use, then determine greatest utility vs risk. From initially working to volume production & implantation is a long road. As with vehicle safety, it will be much harder to pass our internal standards than minimum regulatory standards.”
“@Erdayastronaut @BocaChicaGal @NASASpaceflight Dome to barrel weld made it to 7.1 bar, which is pretty good as ~6 bar is needed for orbital flight. With more precise parts & better welding conditions, we should reach ~8.5 bar, which is the 1.4 factor of safety needed for crewed flight.”
“@realRickyReeves @Erdayastronaut @NASA Crew Dragon is capable of propulsive landing, but would require extensive testing to prove safety. Better to focus on Starship.”
“@kulpability @NASA Crew Dragon should be physically ready & at the Cape in Feb, but completing all safety reviews will probably take a few more months”
“@indiealexh @es2120 @Tesla @woodhaus2 Safety work is done by engineering, with passive safety by body/chassis eng & active by Autopilot. Design in automotive refers to look & feel.”
“@SciGuySpace We had to reallocate some resources to speed this up & received great support from Airborne, our parachute supplier. I was at their Irvine factory with the SpaceX team on Sat and Sun. We’re focusing on the advanced Mk3 chute, which provides highest safety factor for astronauts.”
“@alistairmilne But we probably won’t try for best lap time this week, as we need to review & tune Model S thoroughly for safety on Nürburgring, especially Flugplatz section”
“@TeslaTrip @arctechinc We need to complete internal QA testing for safety, which will hopefully be done in about 4 weeks”
“@HarryStoltz1 @Sofiaan @IIHS_autosafety @atj721 Ultimately, however, what really matters is empirical safety. Nothing in life is 100% safe, so optimal figure of merit is how many millions of miles/km per accident.”
“@HarryStoltz1 @Sofiaan @IIHS_autosafety @atj721 Good question. NHTSA is great for fundamental structural safety, except for unbelted front passenger occupant factor. Very rare these days for front passenger to be unbelted. NCAP also great, but there are some weird headlight angle factors. Super inside baseball stuff …”
“@kulpability @lexfridman @Tesla Intersections with complex traffic lights & shopping mall parking lots are the two biggest software challenges. Developer branch already mostly works in these scenarios, but massive effort required to get to 99.9999% safety.”
